Rapid and precise quantification of lymphocyte iron content by single cell inductively coupled plasma mass spectrometry

20 Metals facilitate catalysis during cellular metabolism, but heterogeneity of metal 21 content at single-cell level within and between cell populations is poorly 22 characterized. This is important because deficiencies of biometals, for example iron, 23 are enormously prevalent worldwide. Here we quantify metal content of single-cells 24 using inductively-coupled plasma mass spectrometry. To develop the method, we 25 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint used rhodium and iridium-intercalated Jurkat cells, obtaining >0.96% r2 cross-26 analytical correlation with mass cytometry. We quantified iron and calcium mass/cell 27 for murine T-lymphocytes with 3% and 8% 2-sigma intra-precision, respectively, 28 when assessing thousands of cells/minute. T-lymphocytes exposed to a 625-fold 29 difference in extracellular iron concentrations maintained close iron homeostatic 30 control, varying ~20% in iron content. Nevertheless, this relatively small variation 31 strongly correlated with changes in cellular activation characteristics measured by 32 flow cytometry. We also assessed human B-cell iron content, which was ~10-fold 33 higher than murine T-lymphocytes. Overall, we demonstrate rapid iron quantification 34 at single-cell level in different cell types and relate cellular iron content to cell 35 function. 36 37 Teaser 38 Precise and rapid iron metallomics of lymphocytes by single cell ICP-MS is a 39 powerful approach for accessing signatures of immunological status. 40 41

42 Iron is crucial for multiple cellular biochemical activities(1). The failure to maintain 43 iron homeostasis has important consequences for human health and vitality(2). For 44 example, hereditary haemochromatosis results in iron accumulation, leading to tissue 45 and organ damage(3–5). Iron deficiency anaemia (IDA) is recognised as the world’s 46 most common micronutrient deficiency, affecting over a billion people worldwide (6). 47 Furthermore, the severity and prevalence of IDA is particularly felt in many low-48 middle income countries, where a combination of malnutrition and endemic diseases 49 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint give rise to unfavourable fates of disease pathogenesis, neurological developments 50 and mortality(7). 51 IDA can also hinder the functioning of the immune system with consequences for 52 health, including vaccine response(8). Stable homeostatic control of iron is crucial for 53 effective functioning of T-cells and B-cells, where iron uptake is essential for cellular 54 proliferation(9, 10), activation(11) and metabolic reprogramming(12). 55 Iron deficiency is typically characterised by measurement of plasma parameters such 56 as ferritin, or transferrin saturation, which act as proxy assessments of iron stores and 57 iron availability for cells, respectively. However, the manifestations of iron deficiency 58 depend on the iron content of cells and how cellular functions are altered by lack of 59 iron. Our knowledge of how iron is distributed between different cell types and how 60 cells respond to different levels of iron availability is currently limited(13). This 61 largely reflects the inherent analytical difficulties in accurately measuring small 62 variations in cellular iron levels amongst representative populations of cells. 63 Currently, cellular iron contents are typically determined by bulk dissolution, 64 followed by analysis of the resultant solutes by a variety of atomic spectroscopy 65 techniques (for example, inductively coupled plasma atomic emission 66 spectrometry(14)). Although appropriate for determination of trace element contents, 67 bulk approaches such as these are incapable of revealing any heterogeneity amongst 68 cell populations. This, coupled with appreciable errors arising from cell counting 69 methods, may lead to large uncertainties in the derived mass per cell values. 70 Analyses of individual cells by microanalytical methods, such as nanoSIMS and 71 synchrotron µXRF, can enable the elemental quantification on a per-cell basis. 72 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint However, these microanalytical approaches are limited by the number of cells that 73 may be analysed, the fixing procedures required and, in the case of synchrotron-based 74 instruments, timely access to the instrument. Although they may deliver an accurate 75 measure of cellular elemental content, the small number of cells that can be 76 realistically analysed hinders estimates of intra-population variation. 77 An alternative approach to quantify the iron content of individual cells is to employ 78 single cell inductively coupled plasma mass spectrometry (SC-ICP-MS). This 79 technique relies on aspirating a solution containing a suspension of cells into an ICP-80 MS, such that they are dispersed sufficiently to be ionised, and subsequently analysed, 81 sequentially. The ICP-MS instrument is operated in time-resolved mode, where 82 analytical time intervals are set sufficiently short that cells transiting through the 83 instrument can be discriminated from the background and ion counts arising are 84 integrated to yield a total cellular mass for the element of interest. The SC-ICP-MS 85 approach allows for the analysis of individual cells at a frequency of hundreds to 86 thousands of cells per minute, sufficient for exploring cell population heterogeneity 87 and providing a potentially rapid screening tool. However, measurement of iron by 88 this analytical method has been successfully described in only a very few research 89 studies. Upon examination of the literature and to our best knowledge, only one study 90 has successfully quantified single cell iron by SC-ICP-MS (specifically in Raji cells 91 for chemotherapy-drug uptake(15)), in addition to two others which have quantified 92 iron levels in bacteria(16, 17). 93 In this study, we optimised SC-ICP-MS for cellular material and cross-validated data 94 against both ‘traditional’ bulk ICP-MS and CyTOF. We assessed single cell 95 metallomics in populations of three different types of lymphocyte (Jurkat cells, OT-I 96 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint murine T-cells and human primary B-cells). We measured iron in murine T-cells 97 exposed to iron sufficient and iron depleted conditions and related iron content to cell 98 activity. 99

100 Single cell metallomics by SC-ICP-MS validated by titrated intercalation 101 experiments. Rhodium and iridium, in their 3+ cationic forms are very effective 102 metallointercalators when complexed with polycyclic aromatic ligands(18, 19). They 103 are often used in high-dimensional phenotypic/ functional single cell analyses by 104 CyTOF mass cytometry for this reason (20). Moreover, as neither of these metals are 105 normally utilised in biology, are both chemically rare (thus unlikely to reveal any 106 contamination issues), and in their ionic form generate highly sensitive signals by 107 ICP-MS, they can also be used to validate SC-ICP-MS as a metrological metallomic 108 analytical technique. 109 In this experiment, titrated concentrations of rhodium and iridium intercalators were 110 added to Jurkat cells, a prototypical T-cell line, and uptake was assessed in the 111 attogram/cell (10-18g) mass range by SC-ICP-MS, Mass Cytomtery (CyTOF) and bulk 112 ICP-MS (see Fig. 1A). PerkinElmer’s Syngistix Single Cell software module was 113 used for realtime viewing of the single cell metallomics and post-measurement 114 evaluation of SC-ICP-MS analysis, where the peak area from each cell event was 115 used to determine the mass of metal per cell. The PerkinElmer NexION 350D ICP-116 MS was used to measure the rhodium-intercalated cells, and the PerkinElmer NexION 117 5000 for the iridium-intercalated cells. 118 The single cell metallomic approach using ICP-MS (as SC-ICP-MS) allows each cell 119 from a population to be sequentially introduced to the instrument, where they are 120 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint individually vapourised, atomised, and their metallomic constituents ionised for 121 measurement within durations of around 200-400 µseconds(21). Known as a “cell 122 event”, it is essential the instrument is able to capture the measurement of the 123 metallomic plume for a particular element during its transit through the mass 124 spectrometer. In this study we harness the high-frequency scanning capability of the 125 PerkinElmer NexION ICP-MS to measure metals as mass-spectral fingerprints with 126 minimal detector exposure times (dwell times), at intervals that are much lower than 127 cell event durations (up to frequencies of 100,000Hz). This enables us to rapidly 128 collect highly precise single cell metallomics, and to measure hundreds of cells per 129 minute. 130 Fig. 1B present excerpts of realtime metallomic data taken from the rhodium and 131 iridium SC-ICP-MS analyses of the intercalated cells, where each individual spike 132 represents a single cell event. The peak insets demonstrate the high level of detail per 133 cell event, where the beginning, end and apex of the example events are clearly 134 defined (see insets within Fig. 1B). In the literature it has been postulated that low 135 detector dwell times can adversely affect SC-ICP-MS data quality(22), primarily from 136 compromised signal intensities. However, from our data we instead found that the 137 detailed peak profiles, in addition to extremely low background baselines, conversely 138 enhanced analytical figures of merit (transferable metrics for analytical performance 139 (23)), such as linearity and precision. Moreover, we prove that high-frequency 140 scanning also provides the prospect of filtering out fragmented cells versus real cell 141 events, where the former transit through the instrument from distinguishably shorter 142 event times than intact cells (see Fig. S1 in supplementary text). 143 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Mass frequency histograms from the rhodium and iridium SC-ICP-MS measurements 144 of each intercalated condition are shown in Figs. 1C and 1G, respectively, which are 145 presented with optimised bin-sizes and log-normal fittings. Background corrections 146 were also applied, which were defined by eliminating all cell events that fell short of 147 the signal duration of a well-defined peak from the mass spectra (see supplementary 148

for further information). Following such corrections, clear successions in 149 metal mass per cell are found with increasing intercalation concentration, revealing 150 the effectiveness of SC-ICP-MS to measure metallomic profiles of entire cell 151 populations throughout 10-18g mass ranges. Heterogeneity of cellular metal uptake is 152 also well-defined within each population, where interquartile ranges of 48ag (Titrant 153 2/ 0.1µmol) to 385ag (Titrant 6/ 0.5 µmol) were found from the Rh intercalation 154 experiment; in addition to 7.5ag (Titrant 2/ 0.01 µmol) to 36ag (Titrant 6/ 0.05 µmol) 155 from the Ir intercalation experiment (see Table S1in supplementary text). 156 Additionally, the linearity in single cell mass progression throughout both titrations is 157 clearly demonstrated from the clear linear regressions, significant r2 values (>0.97) 158 and p-values (0.002) exhibited in Figs. 1D and 1H, where scatter plots of the 159 population geometric means against the titrated intercalator molarity are presented. 160 Duplicate and triplicate aliquots from each condition of both titrations were measured 161 by CyTOF and bulk ICP-MS to compare signal outputs. Excellent agreements were 162 found between SC-ICP-MS and CyTOF from both titrations, which are presented in 163 Figs. 1E and 1I as scatter plots of mean datapoints found from each sample 164 (geometric means of metallomic distributions gained from the SC-ICP-MS and 165 arithmetic means from CyTOF), where r2 >0.96 and p-values ≤0.002 are shown in 166 both correlations. A similar correlation plot comparing SC-ICP-MS to bulk ICP-MS 167 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint is also presented in Fig. 1F from the rhodium intercalation experiment, where on this 168 occasion a much weaker correlation is displayed (r2 = 0.63, p-value = 0.11). 169 The contrast in analytical performance found between the single cell and bulk 170 analytical techniques tested in this experiment is driven by the considerably enhanced 171 detection capability from the former methods, where the combination of better fittings 172 to the linear regression model (revealed from the enhanced r2 values) and the 173 attainment of highly significantly p-values (<0.05) from each titration are found (see 174 Figs. 1D, 1E, 1H and 1I). This is further emphasized by the bulk ICP-MS results 175 gained from the lower-level iridium titration, where most conditions were found 176 below the limit of quantification (thus not presented in Fig. 1). 177 Calcium in T-cells is effective for determining transport efficiency for SC-ICP-178 MS. CD8+ T-cells were extracted from three OT-I mice. Fig. 2A presents an excerpt 179 of the realtime single cell iron output from the T-cells analysed in this experiment, 180 where akin to the previous rhodium and iridium intercalation studies, high frequency 181 scanning was employed to capture high-definition profiles of each single cell ionic 182 plume event. 183 Additionally, we also used SC-ICP-MS to rapidly scan for calcium within the same 184 cell suspensions (see Fig. 2B for an excerpt of realtime calcium metallomic output). 185 In lymphocytes, calcium is an endogenous element that is contained within a similar 186 mass range to iron (circa. one magnitude higher)(24). Coupling the measurement of 187 calcium metallomics to this experiment provided a direct means of determining the 188 transport efficiencies of cell metallomic detection – the proportion of cells entering 189 the plasma and being detected by the ICP-MS over the cell concentration in the 190 measured aliquot, by the particle frequency method(25). Moreover, as the calcium 191 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint mass spectra returned profiles that were constrained within tight mass ranges and 192 similar duration times, it was a very useful proxy for determining a true value of 193 cellular transmission and detection. From these data we determined a mean calcium-194 derived transport efficiency value of 12%, which is within a similar range to other 195 studies who also reported using endogenous elements as a proxy for transport 196 efficiency(15, 26, 27). Furthermore, this accurate representation of cell transmission 197 is also complemented with high precision figures of merit, which was demonstrated 198 by the return of excellent signal to background ratios. This was not only attained by 199 scanning directly on calcium’s major isotope (40Ca) through the coupling of the 200 instrument’s dynamic reaction cell (DRC) and tandem mass spectrometry capability, 201 but also through the utilisation of high purity Maxpar® Cell Acquisition Solution Plus 202 for an extremely low background (see Table 1 and Methods for details). The inset 203 peak within Fig. 2B illustrates this, where signal intensities captured from the apex of 204 the peak from this particular cellular event, were over 30 times higher than the 205 average baseline readings adjacent to the peak [n=17]. 206 The single cell iron data also presents similarly tightly constrained cell event 207 populations (for example as shown in Fig. 2A), but in contrast to the calcium 208 metallomic data sporadic high-mass events were also present in the mass spectra 209 profiles – which were most likely oxidised iron precipitates (as shown in Fig. 2A by 210 the see peaks >200 counts in amplitude at ~4 seconds and ~9 seconds). However, 211 from our high frequency scanning methodology, we were able to differentiate these 212 discrete events from the data by their distinctly higher peak heights and widths. 213 Additionally, as they also lacked any defined gaussian/ lognormal distributions and 214 consisted of values away from the well-defined cell distributions, threshold 215 corrections from the cell distributions was relatively simple. After subsequent 216 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint comparative filtering of the iron quantification data for only cell events, the post-217 filtered iron metallomic data presents a mean transport efficiency that is only 1% 218 higher to the value obtained from calcium, at 13%, thus proving a highly accurate 219 filtering method. Table S2 in supplementary text provides data of the transport 220 efficiency derived for calcium and iron (after filtering) for each sample, in addition to 221 similarly calculated transport efficiencies derived from the measurement of holmium 222 scanning from EQTM Four Element beads. In contrast to cells, these polystyrene beads 223 were resistant to lysing during the introduction phase of measurement, which is 224 highlighted by the higher values, where an average transport efficiency of 27% was 225 attained (see Table S2). 226 Iron analysis by SC-ICP-MS requires chemical resolution by MS/MS for 227 accurate detection. Since its inception in the 1990s, the DRC within PerkinElmer 228 ICP-MS instruments (in addition to other similar manufactured reaction cells), has 229 revolutionised the ability to accurately measure those elements that are affected by 230 polyatomic interference (28, 29). Its ability to negate polyatomic interferences, either 231 by adopting exothermic reactions from reactive gases to either neutralise/ disassemble 232 such ions, or by kinetic energy modulation/ dissociation from highly pressurised inert 233 gases, has elevated analytical performance for a plethora of elements, including iron. 234 However, such gases also sustain impedances (although tempered) to the analyte ions, 235 which can complicate SC-ICP-MS analysis. This effect causes time-elongation of ion 236 plume events and thus peak tailing, up to 6ms in the time-resolved spectra(30), which 237 can detrimentally impact upon the accuracy of the resultant metallomic/ nanoparticle 238 findings(15, 31). Additionally, it can also increase the probability of recording 239 doublets, especially when using high flow rates of NH3 as the cell gas for mass-shift 240 scanning methods(15). We mitigated this impact in our experiments, by using a 241 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint moderated flow of NH3 in the reaction cell, elicited by an MS/MS on-mass scanning 242 approach for analysis – a chemical resolution method that still remains effective at 243 removing such problematic spectral interferences, but one that does not inflict such 244 significant impacts on the ionic energies(15, 30). Here we observed only minimal 245 elongations to such cell event data, where ~1ms peak widths were reported by 246 chemical resolution (iron), versus circa. 0.5ms when no cell gas was used (rhodium) 247 (see Figs. 2A and 1B, respectively). This magnitude of protraction is 6-times lower 248 than those reported from the mass-shift approach(30), engendering a negligible 249 probability of recording doublets (i.e. from the occurrence of two simultaneous peak 250 events). 251 To quantify uncertainty from our measurement method we examined analytical 252 figures of merit, including precision and accuracy. For precision, determinations from 253 consecutive measurement repeats from one of the samples measured [n=4] were 254 evaluated, which revealed 2-sigma variability (as 2 times relative standard deviation) 255 of 2.7% for iron and 8.3% for calcium. Additionally, repeat measurements of EQTM 256 Four Element Calibration Beads for holmium (165Ho) at intervals throughout the 257 analytical run [n=4] presented similarly-calculated precision values of 5.9%. As there 258 are no certified reference materials available for this field of analysis, analytical 259 accuracy was instead determined by measurement of an iron quality-control standard 260 solution. This presented relative errors of 3.1% for iron and 8.1% for calcium. 261 Cellular iron uptake in OT-I T-cells is limited by proliferation. OT-I T-cells 262 derived from three individual mice were activated for 48 hours in iron-free media 263 supplemented with titrated concentrations of holotransferrin (transferrin protein with 264 two iron atoms bound) from 0.001mg/mL to 0.625 mg/mL, which characterised a 265 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint range of conditions from severe iron deficiency to iron replete (see Fig. 3A). Mass 266 frequency histograms of the background corrected single cell iron data for the titrants 267 from each mouse are presented in Fig. 3B, which are presented with optimised bin 268 sizes, log-normal fittings and annotated geometric mean values for each population. 269 We observed only small increments in the geometric means of cellular iron content in 270 populations exposed to increasing amounts of holotransferrin. These extrapolated 271 average iron contents per cell (for each condition and each mouse), were subsequently 272 plotted against each media iron concentration, where clear correlations are presented 273 (see Fig. 3E). Although the geometric means of atomic iron content per cell only 274 increased slightly (~20%) over a 625-fold difference in holotransferrin, this indicates 275 a general maintenance of cellular iron homeostasis in the face of a range of 276 extracellular iron availability. 277 Heterogeneity within the main distributions of each population was evaluated from 278 their calculated interquartile ranges, where similar limited increases between the 279 lowest and highest iron conditions are presented (see Fig. S2 in supplementary text). 280 Equally, an evaluation of intra-population variability, in addition to diversity in iron-281 uptake behaviour at the tails of the populations was also determined upon segregating 282 each population into percentiles, where means of each 10th percentile were calculated 283 and plotted per mouse in Fig. 3C. Between the 10th and 70th percentiles smooth trends 284 are illustrated, reflecting well-constrained and tightly-defined log normal distributions 285 in iron per cell (see Fig. 3B), which also transition in magnitude in accordance to their 286 iron condition. However, beyond the 70th percentiles, significant escalations in the 287 average cellular iron levels are presented in every condition, which is replicated from 288 cells isolated from each mouse. This observation, although characteristic of positively 289 skewed distributions, indicates that the top 20% of cells within each condition contain 290 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint discrete elevations in iron levels over the rest of the population. Furthermore, unlike 291 the uniform transitions noted from the earlier percentiles, the rapid increases in 292 average iron/cell within this range do not occur linearly, where distinctly higher 293 elevations were found for the higher conditions (particularly 0.625mg/mL). 294 Unlike the previous intercalation studies, during the 2-day culturing period for this 295 murine T-cell experiment, live cells were able to continually acquire available iron 296 and proliferate (see Table S3 in supplementary text). To examine this effect, we 297 analysed parallel aliquots of cell cultures by flow cytometry to assess the extent of 298 cell division per condition for each mouse. The proportions of cells within each 299 population that didn’t divide, divided once, or divided two or more times over the 48-300 hour period, are presented in Fig. 3D. The results present clear evidence of the 301 escalation in proliferative activity in the higher iron-bearing conditions, particularly 302 the proportion of cells undergoing 2+ divisions. 303 As mentioned above, although the geometric mean of iron content per cell from this 304 experiment only varied by ~20%, the total amount of cellular iron incorporated into 305 the cell population is higher. Using live cell count data collected during cell harvest 306 together with the geometric means of iron content/cell, it was possible to quantify the 307 total amount of iron used by the cell populations in this experiment (see Fig. 3F). Of 308 significance, the peak iron yield did not associate with the highest condition, which 309 was instead found at the penultimate condition level (0.125mg/mL). This is likely a 310 consequence of cytotoxic effects from the superfluous levels contained within the 311 0.625mg/mL condition, which is reflected by the lower live-cell counts obtained. 312 Furthermore, unlike the low levels of variability found in single cell iron levels 313 between each iron condition, much wider differences were found in the total amounts 314 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint of iron consumed per cell population, where a maximum variability of 94% was 315 found between the highest and lowest calculated values. 316 The extent of proliferative activity was also found to correlate with the shape and 317 tailing of each single cell iron population obtained by SC-ICP-MS, which was 318 measured by their skewness values (see Fig. 3G). The degree of tailing, or skewness, 319 in such distributions is also proportional to the amount of heterogeneity in the cell 320 populations, thus revealing crucial detail on the spread of metallomic masses, or even 321 potentially phenotypical variability. Across the culture conditions from this 322 experiment, the skewness of the single cell iron mass distributions inversely 323 correlated with their degree of proliferation (proportion of cells divided), where those 324 populations comprising elevated skewness also posed enhanced phenotypical 325 variability – affirmed by the associated increase in non-dividing cells that must 326 accumulate higher levels of iron within the population (see Fig. 3G). Furthermore, all 327 of the datapoints from mouse 3, with the exception of one, contain the lowest 328 skewness values, together with the highest levels of cell division, which could explain 329 why lower geometric means of the single cell iron distributions were presented from 330 this mouse (see Figs. 3E and 3G). Also, as the cell division findings provide a high 331 degree of correlation to the single cell iron data from the ICP-MS, we can discount 332 any possibility of the high iron mass per cell datapoints being recorded as doublet cell 333 measurements. 334 Single cell iron metallomic data correlates with surface glycoprotein markers. In 335 addition to measuring cell proliferation activity, flow cytometry was also employed to 336 assess cell surface expression of CD71 (transferrin receptor) and CD25 (IL-2 337 receptor) per condition for each mouse. The data collated from these measurements 338 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint are compared to the geometric means of iron mass from each cell distribution 339 measured by SC-ICP-MS, where such data is presented as scatter plots in Figs. 4A 340 and 4B, respectively. CD71 provides the primary uptake mechanism for transferrin-341 bound iron into eukaryotic cells. Notably, mutations that disable CD71-mediated iron 342 acquisition cause immunodeficiency and impair proliferation of T-cells(32). CD71 is 343 normally highly expressed by activated T-cells, but its synthesis is also regulated by 344 intracellular iron content, with relatively iron deficient cells expressing higher levels 345 of CD71 in order to more efficiently capture any available extracellular iron(12). We 346 observed a clear inverse correlation between geometric mean iron content per cell and 347 mean fluorescence intensity of CD71 (Fig 4A), providing an orthogonal assessment of 348 cellular iron content downstream of cell-intrinsic iron sensing mechanisms. 349 CD25 is the alpha chain component of the IL-2 surface receptor on T-cells. IL-2 350 signalling via CD25 promotes T cell growth and facilitates their differentiation after 351 activation(33). We found that CD25 expression positively correlated with geometric 352 mean iron content, in line with the importance of iron acquisition for cellular 353 activation and growth, as also observed by the correlation of increased cellular iron 354 and cell proliferation observed in Fig. 4B. 355 As a comparison, we plotted expression of surface protein markers to the geometric 356 means of calcium mass from each cell distribution measured by SC-ICP-MS. Figs. 4C 357 and 4D present the associated scatter plots, with no correlations of calcium content 358 with either CD71 or CD25, showing the relative specificity of the relationships 359 between iron content and T-cell activation. 360 Precise single cell iron mass distributions from human primary B-cells. To move 361 beyond murine systems and analyse human cells, we examined primary B-cells 362 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint extracted from the peripheral blood of three healthy donors. After 3 days of in-vitro 363 culturing in R10 media the cells were measured by SC-ICP-MS, using a similar 364

used for the measurement of the previous murine T-cells. B cell purity was 365 assessed through analysis of CD19 surface protein expression. 366 An excerpt of the realtime iron metallomic output are presented in Fig. 5A, where 367 unlike the murine T-cell experiment, peaks comprising wider size ranges were 368 obtained. This observed increase in heterogeneity likely indicates increased 369 phenotypical variability in cellular iron from this cell type. In a similar fashion to the 370 murine T cell analyses, background corrections were applied to remove the presence 371 of ionic and fragmented cell artifacts, with transport efficiencies ranging from 30% to 372 52%. 373 Mass frequency histograms of such corrected data for each donor (D1 – D3) are 374 shown in Fig. 5B, which are presented with optimised bin sizes, log-normal fittings 375 and annotated geometric means for each population. Similar mass ranges were found 376 for each sample, where 2-sigma variability was 0.77fg between the three samples. 377 Additionally, wide distributions are presented, which reflects the wider range in peak 378 sizes noted in Fig. 5A, but extrapolation of the mean values, together with the above 379 uncertainty value presents overall metallomic values that are within range to similarly 380 measured Raji B-cells(15). Moreover, the complimentary CD19 provided 381 confirmation of B cell purity in our measured cell aliquots, where recoveries ranging 382 from 93.6% - 95.8% were reported (see Fig. 5C). 383

384 One of the main advantages of SC-ICP-MS for metallomic analyses is its detection 385 capability, where it is able to quantify metals encapsulated within individual cells 386 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint throughout attogram mass range (10-18g)(34). Additionally, unlike other solution-387 based analyses, dilutions applied to reduce the cell concentration never impact on the 388 signal intensity of the cell events, where instead it may help to reduce the dissolved 389 baseline and thus improve signal: noise ratios. In further attempting to achieve higher 390 counting statistics and reduced data volumes, there is a temptation in many SC-ICP-391 MS studies to collate metallomics from dwell times that exceed the transit time of a 392 cell event (circa. 200-400 µseconds(21)). However, we present crucial findings that 393 instead advocate higher scanning frequencies, particularly at intervals much faster 394 than cell events. Nevertheless, it is important to mention that the ability to scan at 395 frequencies greater than the transit times of cell events may be a limiting factor of the 396 particular ICP-MS instrument being used (35, 36). Several studies have interrogated 397 SC-ICP-MS for the highest achieving dwell times for metallomic analyses(e.g. (22, 398 35)), with the outcome often suggesting the adoption of the shortest available interval 399 time. In addition to negating the possibility of recording doublets, reduced 400 backgrounds were particularly favoured from such measurement intervals. Moreover, 401 in our study we also emphasize the necessity to scan cell events at high frequency 402 intervals to also provide elucidation between “real” cell events and cell fragments/ 403 debris, which in this study, particularly enabled the accurate detection of iron in our 404 murine T-cell experiments. 405 SC-ICP-MS does have its limitations though, notably regarding transport efficiency, 406 where components of the sample introduction system can cause significant cell losses. 407 Nevertheless, through the coupling of a dedicated single cell microflow injection 408 system, we were able to constrain such lysing effects and consistently achieve mean 409 cell-derived transport efficiencies of nearly 15% (see Table S2 in supplementary text). 410 Additionally, the proficient nature of SC-ICP-MS to examine heterogeneity within 411 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint cell populations may not be a requirement for all fields of metallomic research, where 412 bulk approaches providing mean values may otherwise suffice (e.g. (37, 38)). 413 Nevertheless, as we discovered from the method evaluations, inaccuracies from cell 414 counting can inhibit overall accuracy when adopting this approach. 415 In this study we provide a fully-validated evaluation of our SC-ICP-MS method from 416 the intercalation experiments, followed by applied murine T-cell and primary B-cell 417 experiments. Through the careful formulation of titrated rhodium and iridium metal-418 intercalated cells, we present the clear capability of our analytical methodology to 419 precisely quantify cellular metal uptake throughout attogram/cell mass range (as 420 described above and shown in Fig. 1). Heterogeneity throughout the main 421 distributions was also evidently captured and evaluated from their interquartile 422 ranges, where fluent progression from the lower (0.005 µmol – 0.05 µmol) Ir, to the 423 higher (0.05 µmol – 0.5 µmol) Rh intercalation experiments was attained (see Table 424 S1 in supplementary text). 425 From in-depth evaluations of the murine T-cell experiment, not only did we observe 426 equivalence in heterogeneity throughout the main population distributions to the 427 intercalation experiments, but also revealed evidence of phenotypical variability 428 within the uppermost percentiles of each population. Significantly, the magnitude of 429 iron mass per cell within this sub-population also escalated with iron condition, 430 suggesting disparities in metal uptake behaviours associated with iron status. 431 Furthermore, we also correlated the skewness of each distribution to the magnitude of 432 proliferation, where a distinct inverse correlation presented direct evidence of the 433 influence that the magnitude of proliferation had on the overall heterogeneity of the 434 population. The geometric means of each population provided comprehensive 435 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint broader-scale assessments, where moderated uptake profiles were found for each 436 mouse - ascertained from only ~20% increases observed. This occurred even though 437 the media concentrations were titrated throughout a range that varied by a factor of 438 625 times. From this we elucidate the overall influence that metabolic activity played 439 during utilisation of bio-available iron from the live culturing, providing the limited 440 iron per cell uptake profiles that we found. Moreover, this relationship replicates 441 findings from a similar iron culturing study, where magnetotactic bacteria, rather than 442 lymphocytes, were examined for their metabolic response to varying extracellular iron 443 conditions (16). The logarithmic iron uptake trends presented in (16) are analogous to 444 those presented here, which provides encouraging confirmation of the results that we 445 have found. Finally, we provide validation data towards our SC-ICP-MS metallomic 446 findings from glycoprotein markers measured by flow cytometry (CD71 and CD25). 447 Both sets of results provide significant correlations that provide significant direct 448 evidence of metabolic activity to downstream iron contents. 449 Overall, these data findings show the powerful application that SC-ICP-MS can offer 450 in metallomic research, where its ability to rapidly and precisely scan the profiles of 451 entire cell populations can reveal important physiological and/or biochemical 452 behaviours in biological research. Nevertheless, as this is still an emerging field of 453 research there are limitations to the results gained, specifically in the scope of 454

data filtering and the determination of ‘real’ cell results against 455 fragmented entities. As described above we utilise an accurate and robust filtering 456 approach by using the rapid scanning capability of the NexION ICP-MS in addition to 457 ancillary calcium metallomic data to decipher between such data findings, but 458 mathematical models capable of furthering such data corrections are required to 459 reinforce such corrections. 460 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint

461 Jurkat cell line, rhodium and iridium intercalation. Clone E6-1 Jurkats were 462 purchased from ATCC; a clone of the Jurkat-FHCRC cell line (derivative of the 463 original Jurkat cell line). They were cultured in R10 media (RPMI 1640 supplemented 464 with 10% foetal bovine serum, 1% penicillin-streptomycin and 1% glutamine) and 465 incubated at 37°C and 5% CO2 in T75 flasks. For the metal intercalation experiments, 466 freshly passaged cells were divided into subsets (each containing cell concentrations 467 of circa. 4x106 cells/mL) and rinsed by centrifugation with Standard Biotools 468 Maxpar® phosphate buffered saline (PBS), prior to subsequent fixation in 4% 469 paraformaldehyde for 10 minutes at room temperature. Titrated concentrations of 470 either rhodium or iridium intercalators (Standard Biotools’ 500µM rhodium Cell-471 IDTM or 125µM iridium Cell-IDTM, respectively) were then doped into each sample to 472 form intercalation concentration ranges of 0.05 µM – 0.5 µM and 0.005µM – 0.05 473 µM, respectively. The cell samples were then stored overnight at 4°C to ensure 474 complete penetration of the Cell-IDTM organo-metallic compounds by passive 475 diffusion through the permeated membranes of each cell. The following day each 476 sample was divided into two to provide aliquots for both SC-ICP-MS and CyTOF 477 analysis. Prior to analysis the cells were rinsed with Standard Biotools Maxpar® cell 478 staining buffer by centrifugation to remove any excess metal accumulation from the 479 cell surfaces. 480 Mice and T cell isolation from peripheral blood, iron titration and in-vitro 481 proliferation. OT-I mice (2 x 12-week-old males, and 1 x 13-week-old male), were 482 originally obtained from Audrey Gerard, University of Oxford, and were housed in 483 individually ventilated cages. All animal work was completed under the authority of 484 UK home office project and personal licenses under the Animals (Scientific 485 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Procedures) Act (ASPA) 1986. Mice were sacrificed via rising concentration of CO2 486 followed by cervical dislocation. Plates for CD8+ T-cell culture were pre-treated with 487 5 ug/mL α-CD3 (Biolegend, 100239) in phosphate buffered saline for 2 hours at 488 37°C. Spleen and lymph nodes were collected from euthanised mice and macerated 489 through 40 μm filters using PBS supplemented with 2% fetal bovine serum and 1 mM 490 EDTA (Invitrogen, AM9260G). CD8+ T-cells were isolated from the single cell 491 suspension using the EasySep Mouse CD8+ T-cell isolation kit (Stem Cell 492 Technologies, 19853) and the EasyEights EasySep magnet (Stem Cell Technologies, 493 18103). Isolated cells were stained with cell trace violet (CTV, Invitrogen, C34557) 494 for 8 minutes at 37°C in PBS and then washed. CD8+ T-cells were plated at a 495 concentration of 0.5x106 cells/mL on the α-CD3 pre-treated plates. Cells were grown 496 in iron free media (RPMI1640 (Gibco, 21875034), 10% iron free serum substitute 497 (Pan Biotech, P04-95080), 1% glutamine (Sigma Aldrich, G7513-100ML) and 1% 498 penicillin/streptomycin (Sigma Aldrich, P0781-100ML)) supplemented with set 499 concentrations of holo and apotransferrin. Human holotransferrin (R&D systems, 500 2914-HT-001G) was added at concentrations of 0.001 mg/mL to 0.625 mg/mL. Total 501 transferrin levels were kept at a constant concentration of 1.2 mg/mL by adding the 502 appropriate amount of human apotransferrin (R&D systems, 3188-AT-001G). Cells 503 were also treated with 50 μM β-mercaptoethanol (BME, Gibco, 31350-010), 1 μg/mL 504 α-CD28 (Biolegend, 102115) and 50 U/mL IL-2 (Biolegend, 575402) to activate the 505 cells. CD8+ T-cells were cultured at 37°C, 5% CO2 for 48h. 506 After incubation the cells were harvested, counted and aliquots divided between SC-507 ICP-MS and Flow Cytometry; where ~ 2x106 cells/mL were retained for SC-ICP-MS. 508 The SC-ICP-MS cell aliquots were then rinsed twice by centrifugation with Standard 509 Biotools Maxpar® PBS, followed by resuspension in 1mL 4% paraformaldehyde for 510 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint fixation at room temperature for 10 minutes. The samples were then rinsed twice by 511 centrifugation with Standard Biotools Maxpar® Cell Staining Buffer to remove any 512 excess iron remaining from the cell surfaces, followed by resuspension in Standard 513 Biotools Maxpar® Fix and Perm reagent for overnight storage at 4°C. 514 Human blood donations: B cell isolation, purification and in-vitro proliferation. 515 Blood samples taken from three healthy donors from the John Radcliffe Hospital, 516 Oxford, United Kingdom, were utilised for single cell iron analysis in B-cells by SC-517 ICP-MS. Each sample was collected after obtaining written consent and ethical 518 approval from the University of Oxford’s Central University Research Ethics 519 Committee (CUREC). The samples were collected in EDTA, which was followed by 520 PBMC isolation by density gradient centrifugation: Greiner Bio-One Leucosep tubes, 521 containing 15mL of Lymphoprep (Stem Cell Technologies) and collected blood were 522 centrifuged at 1000 x g for 1 minute at ambient temperature. EDTA blood was 523 extracted into the upper chamber of the Leucosep tube and centrifuged at 1000 x g for 524 15 minutes with no brake. The cloudy buffy-coat layer, containing PBMCs was 525 extracted and the cells were subsequently rinsed twice with R0 media (RPMI 1640 526 supplemented with 1% penicillin-streptomycin and 1% glutamine) and PBS, 527 respectively. CellTrace Violet (Thermo Fisher Scientific) was added to the rinsed 528 PBMCs as a tracer for proliferation and incubated at 37°C in 5% CO2 for 8 minutes. 529 After incubation the cells were rinsed with R10 media, counted and then diluted to 530 8x106 cells per mL in R10 media. Two million cells were subsequently added per well 531 into a 24-well rounded bottom plate, together with aliquots of 0.25mL of R10 media 532 and 0.5mL of R10 media supplemented with 1µg/mL R848 (Stem Cell Technologies) 533 and 10 ng/mL of recombinant IL-2 (PeproTech). The prepared cells were then 534 cultured for 3 days at 37°C in 5% CO2. Following polyclonal stimulation, the cells 535 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint were harvested, washed in R10 media and counted. B-cells were then purified from 536 the other harvested PBMCs by negative selection using a Human B-cell Isolation Kit 537 (Stem Cell Technologies) according to manufacturer’s instructions. Following 538 purification, the B-cells were transferred to a 96-well rounded bottom plate and rinsed 539 with PBS, following Fc Receptor (FCR) blocking and live/dead staining. This was 540 followed by the subsequent labelling of the cells with combinations of anti-CD19-541 PerCP Cy5.5, anti-CD21 (Alexa Fluor 700), anti-CD27 (PE-Cy7), anti-CD38 542 (BV510), anti-CD69 (BV605), and anti-CD71 (PE/Dazzle 594) in PBS in addition to 543 incubation for 20 minutes on water ice and fixation buffer (Biolegend). Prior to 544 intracellular staining with anti-IgG (BV711) and anti-IgD (FITC) the cells were 545 permeabilised with perm buffer for 20 minutes on water ice. Prior to measurement by 546 Flow Cytometry, fluorescence minus one controls (FMOs) were included for each 547 marker, in addition to an unstained control 548 Single cell Inductively coupled plasma mass spectrometry (SC-ICP-MS). 549 Following all of the methodologies described above, the prepared cell suspensions 550 were also rinsed a further three times in Standard Biotools’ Maxpar® Cell Acquisition 551 Solution Plus (CAS+) prior to SC-ICP-MS analysis. This was undertaken to ensure 552 both the removal of any remaining residually-retained metals from the cell surfaces, 553 in addition to an exchange into a suspension media suitable for analysis by this 554 technique. Indeed, this reagent is proven to be an optimal choice for analysis utilising 555 our analytical setup over other commonly used carrier reagents(39, 40), where its 556 combination of a neutral pH in addition to a higher ionic content than water provides 557 a higher analytical performance for metallomic analysis when combined with a wider 558 bore injector. Succeeding this rinsing protocol, the final suspensions were filtered 559 through 35µm nylon mesh filters, and the subsequent cell suspensions counted, 560 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint diluted to 105 cells/mL cell concentrations (if required) and then immediately 561 measured by SC-ICP-MS. Supernatants from the final rinse cycles of selected 562 samples were also retained for measurement, to test for any leakage of intracellular 563 metals. 564 All SC-ICP-MS measurements were conducted using either a NexION 5000 multi-565 quadrupole ICP-MS (PerkinElmer) or a NexION 350D ICP-MS (PerkinElmer), in 566 time-resolved mode (see instrument conditions stated in Table 1). Both instruments 567 were equipped with an Elemental Scientific Inc. single cell introduction system, 568 which comprised of a CytoNeb 50 nebuliser, a CytoSpray linear pass spray chamber, 569 a 2.0mm tapered injector (PerkinElmer White Cassette torch with 2.0mm injector for 570 the NexION 5000) and a microFAST autosampler (which provided an additional final 571 agitation of the suspension using its ‘Mix’ submethod to ensure homogenisation of 572 the aliquoted cell suspension for analysis). This apparatus, like many others also used 573 for single cell metallomic research(15, 22, 41), was essential for this analysis to 574 ensure the highest levels of cell transmission to the instrument, where micro-flow 575 volume injections of cells were analysed for precise single cell metallomics by the 576 mass spectrometer from discrete measurements of their resultant ionic plumes. Details 577 relating to the optimisation of this instrument setup is described in supplementary 578 information. 579 Bulk Inductively coupled plasma mass spectrometry (bulk-ICP-MS). Bulk ICP-580 MS analysis utilised cell aliquots remaining after the metal intercalation SC-ICP-MS 581 analyses, where approximately 0.2x106 cells were dissolved in 2% v/v HNO3 within 582 metal-free centrifuge tubes at ambient temperature for 48 hours. Measurements were 583 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint conducted using the NexION 350D ICP-MS (PerkinElmer) using instrument 584 conditions described in Table 2. 585 Mass Cytometry (CyTOF). Mass Cytometry was used in this study to validate 586 metallomic data obtained by SC-ICP-MS from the metal intercalation experiments on 587 Jurkat cells (as described above). In an analogous approach to the SC-ICP-MS 588 analyses, the prepared cell subsets allocated for Mass Cytometry measurements were 589 similarly rinsed three times by centrifugation with CAS+ to fully removal any 590 surface-retained metals and to transfer into the instrument carrier media. All 591 measurements by this technique were conducted using a Standard Biotools Helios 592 CyTOF, which employed the use of its standard single cell suspension pneumatic 593 sample introduction system. The mass cytometer was tuned and its performance 594 confirmed using EQTM Four Element Calibration Beads. The cells were diluted to 106 595 cells/mL in CAS+ with 10% EQTM Four Element Calibration Beads. The .fcs files 596 were acquired and then processed, including normalisation in CyTOF Software v.7 597 (Standard Biotools). 598 Flow Cytometry. Flow Cytometry analysis was incorporated into this study to 599 validate the iron metallomic data obtained by SC-ICP-MS. Cells were transferred to 600 96 well round bottom plates and washed with PBS. Cells were stained with a cocktail 601 of antibodies and the Zombie NIR fixable viability kit (1:1000, Biolegend, 423105) 602 prepared in PBS for 20 minutes on ice. Cells were subsequently fixed with 2% 603 paraformaldehyde (Pierce, 28906) for 20 minutes on ice, washed and resuspended in 604 PBS. Cells were analysed on either a BD Biosciences LSR FortessaTM X50 605 instrument or a Attune NxT flow cytometer (Thermofisher Scientific). Data was 606 analysed using FlowJo (BD biosciences).. 607 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Statistical Analysis. All individual single cell metallomic mass data was calculated 608 using the single cell module within PerkinElmer’s SyngistixTM ICP-MS software. 609 Complimentary geometric means and p values were calculated in Excel, where the 610 latter was determined using the ‘Regression’ data analysis toolpack. Additionally, 611 other linear regression determinations, such as linear fitting equations and r2 values 612 were calculated in R, utilising the dplyr and ggplot2 libraries. 613

614 615 1. M. R. Teh, A. E. Armitage, H. Drakesmith, Why cells need iron: a compendium 616 of iron utilisation. Trends in Endocrinology & Metabolism, 617 S1043276024001097 (2024). 618 2. P. T. Lieu, M. Heiskala, P. A. Peterson, Y. Yang, The roles of iron in health and 619 disease. Molecular Aspects of Medicine 22, 1–87 (2001). 620 3. R. Acton, J. Barton, L. Passmore, P. Adams, G. Mclaren, C. Leiendeckerfoster, 621 M. Speechley, E. Harris, O. Castro, J. Reiss, Accuracy of Family History of 622 Hemochromatosis or Iron Overload: The Hemochromatosis and Iron Overload 623 Screening Study. Clinical Gastroenterology and Hepatology 6, 934–938 (2008). 624 4. L. Sousa, M. M. Oliveira, M. T. C. Pessôa, L. A. Barbosa, Iron overload: Effects 625 on cellular biochemistry. Clinica Chimica Acta 504, 180–189 (2020). 626 5. C. Thomas, M. M. Mackey, A. A. Diaz, D. P. Cox, Hydroxyl radical is produced 627 via the Fenton reaction in submitochondrial particles under oxidative stress: 628 implications for diseases associated with iron accumulation. Redox Report 14, 629 102–108 (2009). 630 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint 6. S.-R. Pasricha, J. Tye-Din, M. U. Muckenthaler, D. W. Swinkels, Iron 631 deficiency. The Lancet 397, 233–248 (2021). 632 7. A. M. Prentice, H. Ghattas, C. Doherty, S. E. Cox, Iron metabolism and malaria. 633 Food Nutr Bull 28, S524-539 (2007). 634 8. A.-M. Mitterstiller, L. Von Raffay, M. Nairz, “Iron Deficiency, Anemia, and the 635 Immune System” in Nutritional Anemia, C. D. Karakochuk, M. B. 636 Zimmermann, D. Moretti, K. Kraemer, Eds. (Springer International Publishing, 637 Cham, 2022; https://link.springer.com/10.1007/978-3-031-14521-6_18)Nutrition 638 and Health, pp. 235–248. 639 9. Y. Jiang, C. Li, Q. Wu, P. An, L. Huang, J. Wang, C. Chen, X. Chen, F. Zhang, 640 L. Ma, S. Liu, H. He, S. Xie, Y. Sun, H. Liu, Y. Zhan, Y. Tao, Z. Liu, X. Sun, Y. 641 Hu, Q. Wang, D. Ye, J. Zhang, S. Zou, Y. Wang, G. Wei, Y. Liu, Y. Shi, Y. 642 Eugene Chin, Y. Hao, F. Wang, X. Zhang, Iron-dependent histone 3 lysine 9 643 demethylation controls B cell proliferation and humoral immune responses. Nat 644 Commun 10, 2935 (2019). 645 10. S. Ni, Y. Yuan, Y. Kuang, X. Li, Iron Metabolism and Immune Regulation. 646 Front Immunol 13, 816282 (2022). 647 11. F. J. Hartmann, D. Mrdjen, E. McCaffrey, D. R. Glass, N. F. Greenwald, A. 648 Bharadwaj, Z. Khair, S. G. S. Verberk, A. Baranski, R. Baskar, W. Graf, D. Van 649 Valen, J. Van Den Bossche, M. Angelo, S. C. Bendall, Single-cell metabolic 650 profiling of human cytotoxic T cells. Nat Biotechnol 39, 186–197 (2021). 651 12. J. N. Frost, T. K. Tan, M. Abbas, S. K. Wideman, M. Bonadonna, N. U. Stoffel, 652 K. Wray, B. Kronsteiner, G. Smits, D. R. Campagna, T. L. Duarte, J. M. Lopes, 653 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint A. Shah, A. E. Armitage, J. Arezes, P. J. Lim, A. E. Preston, D. Ahern, M. Teh, 654 C. Naylor, M. Salio, U. Gileadi, S. C. Andrews, S. J. Dunachie, M. B. 655 Zimmermann, F. R. M. van der Klis, V. Cerundolo, O. Bannard, S. J. Draper, A. 656 R. M. Townsend, B. Galy, M. D. Fleming, M. C. Lewis, H. Drakesmith, 657 Hepcidin-Mediated Hypoferremia Disrupts Immune Responses to Vaccination 658 and Infection. Med 2, 164-179.e12 (2021). 659 13. L. Galera-Laporta, C. J. Comerci, J. Garcia-Ojalvo, G. M. Süel, IonoBiology: 660 The functional dynamics of the intracellular metallome, with lessons from 661 bacteria. Cell Systems 12, 497–508 (2021). 662 14. C. Ekmekcioglu, C. Prohaska, K. Pomazal, I. Steffan, G. Schernthaner, W. 663 Marktl, Concentrations of Seven Trace Elements in Different Hematological 664 Matrices in Patients with Type 2 Diabetes as Compared to Healthy Controls. 665 BTER 79, 205–219 (2001). 666 15. T. Liu, E. Bolea-Fernandez, C. Mangodt, O. De Wever, F. Vanhaecke, Single-667 event tandem ICP-mass spectrometry for the quantification of chemotherapeutic 668 drug-derived Pt and endogenous elements in individual human cells. Analytica 669 Chimica Acta 1177, 338797 (2021). 670 16. M. Amor, M. Tharaud, A. Gélabert, A. Komeili, Single‐cell determination of 671 iron content in magnetotactic bacteria: implications for the iron biogeochemical 672 cycle. Environ Microbiol 22, 823–831 (2020). 673 17. M. Ikehata, J. Woolcock, M. E. P. Murphy, “Iron Content Measurement in 674 Individual Bacterial Cells Using SC-ICP-MS” (Application Note, PerkinElmer, 675 Inc., 940 Winter Street, Waltham, MA 02451, USA, 2018); 676 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint https://resources.perkinelmer.com/lab-677 solutions/resources/docs/app_013978_01_nexion_sc-icp-678 ms_iron_in_bacterial_cells.pdf. 679 18. H. Li, Q. Yang, Z. Wang, F. Li, Iridium Complex with Specific Intercalation in 680 the G-Quadruplex: A Phosphorescence and Electrochemiluminescence Dual-681 Mode Homogeneous Biosensor for Enzyme-Free and Label-Free Detection of 682 MicroRNA. ACS Sens. 8, 1529–1535 (2023). 683 19. B. M. Zeglis, V. C. Pierre, J. K. Barton, Metallo-intercalators and metallo-684 insertors. Chem. Commun., 4565 (2007). 685 20. V. D. Gonzalez, Y.-W. Huang, W. J. Fantl, Mass Cytometry for the 686 Characterization of Individual Cell Types in Ovarian Solid Tumors. Methods 687 Mol Biol 2424, 59–94 (2022). 688 21. S. D. Tanner, D. R. Bandura, O. Ornatsky, V. I. Baranov, M. Nitz, M. A. 689 Winnik, Flow cytometer with mass spectrometer detection for massively 690 multiplexed single-cell biomarker assay. Pure and Applied Chemistry 80, 2627–691 2641 (2008). 692 22. Y. Cao, J. Feng, L. Tang, C. Yu, G. Mo, B. Deng, A highly efficient introduction 693 system for single cell- ICP-MS and its application to detection of copper in 694 single human red blood cells. Talanta 206, 120174 (2020). 695 23. A. C. Olivieri, G. M. Escandar, “Analytical Figures of Merit” in Practical 696 Three-Way Calibration (Elsevier, 2014; 697 https://linkinghub.elsevier.com/retrieve/pii/B9780124104082000065), pp. 93–698 107. 699 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint 24. A. Kubala-Kukuś, D. Banaś, J. Braziewicz, I. Stabrawa, S. Góźdź, Studies of 700 Element Concentration in the Lymphocytes, Erythrocytes, and Plasma of 701 Healthy Human Donors Using Total Reflection X-ray Fluorescence. Appl 702 Spectrosc 75, 802–817 (2021). 703 25. H. E. Pace, N. J. Rogers, C. Jarolimek, V. A. Coleman, C. P. Higgins, J. F. 704 Ranville, Determining transport efficiency for the purpose of counting and sizing 705 nanoparticles via single particle inductively coupled plasma mass spectrometry. 706 Anal Chem 83, 9361–9369 (2011). 707 26. Y. Tanaka, R. Iida, S. Takada, T. Kubota, M. Yamanaka, N. Sugiyama, Y. 708 Abdelnour, Y. Ogra, Quantitative Elemental Analysis of a Single Cell by Using 709 Inductively Coupled Plasma‐Mass Spectrometry in Fast Time‐Resolved 710 Analysis Mode. ChemBioChem 21, 3266–3272 (2020). 711 27. M. Corte Rodríguez, R. Álvarez-Fernández García, E. Blanco, J. Bettmer, M. 712 Montes-Bayón, Quantitative Evaluation of Cisplatin Uptake in Sensitive and 713 Resistant Individual Cells by Single-Cell ICP-MS (SC-ICP-MS). Anal. Chem. 714 89, 11491–11497 (2017). 715 28. S. D’Ilio, N. Violante, C. Majorani, F. Petrucci, Dynamic reaction cell ICP-MS 716 for determination of total As, Cr, Se and V in complex matrices: Still a 717 challenge? A review. Analytica Chimica Acta 698, 6–13 (2011). 718 29. S. D. Tanner, V. I. Baranov, D. R. Bandura, Reaction cells and collision cells for 719 ICP-MS: a tutorial review. Spectrochimica Acta Part B: Atomic Spectroscopy 720 57, 1361–1452 (2002). 721 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint 30. E. Bolea-Fernandez, D. Leite, A. Rua-Ibarz, T. Liu, G. Woods, M. Aramendia, 722 M. Resano, F. Vanhaecke, On the effect of using collision/reaction cell (CRC) 723 technology in single-particle ICP-mass spectrometry (SP-ICP-MS). Analytica 724 Chimica Acta 1077, 95–106 (2019). 725 31. I. Kálomista, A. Kéri, G. Galbács, On the applicability and performance of the 726 single particle ICP-MS nano-dispersion characterization method in cases 727 complicated by spectral interferences. J. Anal. At. Spectrom. 31, 1112–1122 728 (2016). 729 32. H. H. Jabara, S. E. Boyden, J. Chou, N. Ramesh, M. J. Massaad, H. Benson, W. 730 Bainter, D. Fraulino, F. Rahimov, C. Sieff, Z.-J. Liu, S. H. Alshemmari, B. K. 731 Al-Ramadi, H. Al-Dhekri, R. Arnaout, M. Abu-Shukair, A. Vatsayan, E. Silver, 732 S. Ahuja, E. G. Davies, M. Sola-Visner, T. K. Ohsumi, N. C. Andrews, L. D. 733 Notarangelo, M. D. Fleming, W. Al-Herz, L. M. Kunkel, R. S. Geha, A missense 734 mutation in TFRC, encoding transferrin receptor 1, causes combined 735 immunodeficiency. Nat Genet 48, 74–78 (2016). 736 33. W. Liao, J.-X. Lin, W. J. Leonard, IL-2 family cytokines: new insights into the 737 complex roles of IL-2 as a broad regulator of T helper cell differentiation. Curr 738 Opin Immunol 23, 598–604 (2011). 739 34. C. Stephen, L. Amable, R. Merrifield, Single-Cell ICP-MS Analysis: 740 Quantifying the Metal Concentration of Unicellular Organisms at the Cellular 741 Level. Spectroscopy 33, 33–39 (2018). 742 35. C. Davison, J. Pascoe, M. Bailey, D. J. V. Beste, M. Felipe-Sotelo, Single cell–743 inductively coupled plasma–mass spectrometry (SC-ICP-MS) reveals metallic 744 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint heterogeneity in a macrophage model of infectious diseases. Anal Bioanal 745 Chem, doi: 10.1007/s00216-024-05592-3 (2024). 746 36. K. N. D. S. Leal, A. B. Santos Da Silva, A. Z. B. Aragão, C. H. I. Ramos, A. J. 747 Stewart, M. A. Z. Arruda, Optimizing the performance of single-cell ICP-748 MS/MS for Fe and Zn determination in human umbilical vascular endothelial 749 cells. Microchemical Journal 202, 110696 (2024). 750 37. N. D. Jhurry, M. Chakrabarti, S. P. McCormick, G. P. Holmes-Hampton, P. A. 751 Lindahl, Biophysical Investigation of the Ironome of Human Jurkat Cells and 752 Mitochondria. Biochemistry 51, 5276–5284 (2012). 753 38. I.-L. Hsiao, F. S. Bierkandt, P. Reichardt, A. Luch, Y.-J. Huang, N. Jakubowski, 754 J. Tentschert, A. Haase, Quantification and visualization of cellular uptake of 755 TiO2 and Ag nanoparticles: comparison of different ICP-MS techniques. J 756 Nanobiotechnol 14, 50 (2016). 757 39. B. H. Lee, G. Kelly, S. Bradford, M. Davila, X. V. Guo, E. D. Amir, E. M. 758 Thrash, M. D. Solga, J. Lannigan, B. Sellers, J. Candia, J. Tsang, R. R. 759 Montgomery, S. J. Tamaki, T. K. Sigdel, M. M. Sarwal, L. L. Lanier, Y. Tian, C. 760 Kim, D. Hinz, B. Peters, A. Sette, A. H. Rahman, A Modified Injector and 761 Sample Acquisition Protocol Can Improve Data Quality and Reduce Inter‐762 Instrument Variability of the Helios Mass Cytometer. Cytometry Pt A 95, 1019–763 1030 (2019). 764 40. A. Iyer, A. A. J. Hamers, A. B. Pillai, CyTOF® for the Masses. Front Immunol 765 13, 815828 (2022). 766 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint 41. S. Miyashita, A. S. Groombridge, S. Fujii, A. Minoda, A. Takatsu, A. Hioki, K. 767 Chiba, K. Inagaki, Highly efficient single-cell analysis of microbial cells by 768 time-resolved inductively coupled plasma mass spectrometry. J. Anal. At. 769 Spectrom. 29, 1598–1606 (2014). 770 Acknowledgments and Funding: 771 PH thanks Human Iron Research in Oxford (HIRO), in addition to financial support 772 from PerkinElmer for the funding towards his DPhil. The authors would like to thank 773 Elemental Scientific Inc. for their technical support in commencing analysis with the 774 single cell sample introduction system. 775 776 Author contributions: 777 Conceptualization: HD, JW, MT 778 Methodology: PH, DP, MT, MM, GP 779 Investigation: PH, MT, MM, HC, GP 780 Visualization: PH, GP 781 Supervision: JW, HD, DP 782 Writing—original draft: PH, JW 783 Writing—review & editing: PH, RH, DP, GP, MT, MM, HD 784 785 Competing interests: The authors declare that they have no competing interests. 786 787 Data and materials availability: Tabulated data accompanying the intercalation and 788 murine T-cell experiments are included in the supplementary data at the end of this 789 manuscript. Please contact the corresponding author for further information about the 790 data presented in this manuscript. 791 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Figures and Tables 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 Fig. 1. Validation of SC-ICP-MS from rhodium (blue) and iridium (red) 830 intercalation experiments. (A). Experimental design for validation of SC-831 ICP-MS by a metal intercalation approach; (B). Excerpts of time-resolved 832 mass spectra of intercalated Jurkat cells, intercalated with 0.3µmol rhodium 833 (left) and 0.04µmol iridium (right). Each spike represents a metallomic event, 834 where their individual areas correlate to mass of rhodium or iridium within 835 each measured cell. The metallomic plumes from each metal take ~0.0004 – 836 0.0005s to transit through the instrument (shown in the inset peaks). The inset 837 peaks also illustrate the resolutions achieved for the event profiles, where 838 measurements were taken at intervals of 75µseconds (rhodium) and 839 10µseconds (iridium); (C). Mass-frequency histograms presenting single cell 840 y = 22 + 1.4 ⋅ x, r2 = 0.968y = 22 + 1.4 ⋅ x, r2 = 0.968y = 22 + 1.4 ⋅ x, r2 = 0.968y = 22 + 1.4 ⋅ x, r2 = 0.968y = 22 + 1.4 ⋅ x, r2 = 0.968 ppppp = 0.002= 0.002= 0.002= 0.002= 0.002 0 250 500 750 1 000 0 200 400 600 103Rh CyTOF (mean int.) SC−ICP−MS Rh (ag) E r2 = 0.629r2 = 0.629r2 = 0.629r2 = 0.629r2 = 0.629 ppppp = 0.11= 0.11= 0.11= 0.11= 0.11 0 250 500 750 1 000 0 250 500 750 1 000 Bulk ICP−MS Rh (ag) SC−ICP−MS Rh (ag) F A GM = 129ag GM = 158ag GM = 219ag GM = 502ag GM = 580ag GM = 760ag Titrant 6 (0.5 µmol) Titrant 5 (0.4 µmol) Titrant 4 (0.3 µmol) Titrant 3 (0.2 µmol) Titrant 2 (0.1 µmol) Titrant 1 (0.05 µmol) 0 1000 2000 3000 0.0 0.5 1.0 1.5 2.0 0.0 2.5 5.0 7.5 10.0 12.5 0 10 20 0 20 40 60 0 10 20 30 0 10 20 Mass Rh (ag) Frequency C y = −28 + 1577 ⋅ x, r2 = 0.974y = −28 + 1577 ⋅ x, r2 = 0.974y = −28 + 1577 ⋅ x, r2 = 0.974y = −28 + 1577 ⋅ x, r2 = 0.974y = −28 + 1577 ⋅ x, r2 = 0.974 ppppp = 0.002= 0.002= 0.002= 0.002= 0.002 0 250 500 750 1 000 0.1 0.2 0.3 0.4 0.5 µmol Rh SC−ICP−MS Rh (ag) D GM = 44ag GM = 26ag GM = 28ag GM = 44ag GM = 54ag GM = 64ag Titrant 6 (0.05 µmol) Titrant 5 (0.04 µmol) Titrant 4 (0.03 µmol) Titrant 3 (0.02 µmol) Titrant 2 (0.01 µmol) Titrant 1 (0.005 µmol) 0 100 200 0 1 2 3 4 0 2 4 6 0 10 20 30 0 10 20 30 0 10 20 30 0 10 20 30 Mass Ir (ag) Frequency G y = 12 + 1033 ⋅ x, r2 = 0.97y = 12 + 1033 ⋅ x, r2 = 0.97y = 12 + 1033 ⋅ x, r2 = 0.97y = 12 + 1033 ⋅ x, r2 = 0.97y = 12 + 1033 ⋅ x, r2 = 0.97 ppppp= 0.002= 0.002= 0.002= 0.002= 0.002 0 20 40 60 80 0.01 0.02 0.03 0.04 0.05 µmol Ir SC−ICP−MS Ir (ag) H y = 15 + 0.1 ⋅ x, r2 = 0.984y = 15 + 0.1 ⋅ x, r2 = 0.984y = 15 + 0.1 ⋅ x, r2 = 0.984y = 15 + 0.1 ⋅ x, r2 = 0.984y = 15 + 0.1 ⋅ x, r2 = 0.984 ppppp = 0.0009= 0.0009= 0.0009= 0.0009= 0.0009 0 20 40 60 80 0 100 200 300 400 500 193Ir CyTOF (mean int.) SC−ICP−MS Ir (ag) I 0.005 - 0.05 μmol Ir0.005 - 0.05 μmol Ir0.05 - 0.5 μmol Rh0.05 - 0.5 μmol RhIntercalator additionsIntercalator additions JurkatJurkatcellscells SC-ICP-MSSC-ICP-MS Bulk ICP-MSBulk ICP-MS MassMasscytometrycytometry 12 hours12 hours preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint metallomic distributions from the rhodium intercalation series (GM = 841 population geometric means). Mass quantified in attograms (ag) (10-18g); (D-842 F). Correlation plots of SC-ICP-MS (geometric means from rhodium 843 experiment) versus intercalation molarity, mean CyTOF intensities and bulk 844 ICP-MS analysis, respectively; (G). Mass-frequency histograms presenting 845 single cell metallomic distributions from the iridium intercalation series (GM 846 = population geometric means); (H-I). Correlation plots of SC-ICP-MS 847 (geometric means from the iridium experiment) versus intercalation molarity 848 and mean CyTOF intensities, respectively. Error bars shown in all correlation 849 plots present 2-sigma precision for all SC-ICP-MS data points [n=4]. Error 850 bars from bulk ICP-MS comprise a combination of ICP-MS and cell counting 851 uncertainties, which are combined using the product rule. 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 Fig. 2. Time-resolved single cell mass spectra measured by SC-ICP-MS from 869 murine T-cell experiments. (A). An excerpt of time-resolved mass spectra 870 from iron SC-ICP-MS analysis of T-cells taken from mouse 3 and the 871 0.005mg/mL condition; (B). An excerpt of time-resolved mass spectra from 872 calcium SC-ICP-MS analysis of T-cells taken from mouse 3 and the 873 0.025mg/mL condition. Each spike in both (A) and (B) represents an 874 individual metallomic event, where their individual areas correlate to mass of 875 calcium or iron within each measured cell. The metallomic plumes from each 876 metallomic event take ~0.001s to transit through the instrument (shown in the 877 inset peaks in both (A) and (B)). The inset peaks also illustrate the resolutions 878 achieved for the event profiles, where measurements were taken at intervals of 879 40µseconds (calcium) and 50µseconds (iron). All measurements conducted 880 using the PerkinElmer NexION 5000 ICP-MS. 881 882 A B preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 Fig. 3. Results from murine T-cell iron deprivation experiment. (A). Experimental 921 design for the assessment of iron in murine T-cells that were exposed to iron 922 media conditions ranging from 0.001mg/mL – 0.625mg/mL holotransferrin; 923 (B). Histograms showing the mass-frequency distributions of iron mass/cell 924 for each holotransferrin condition from cells taken from each of three mice. 925 Mass distributions are shown in attograms (ag) (10-18g); (C). Line plots 926 showing the mean iron mass/cell per tenth percentile for each holotransferrin 927 condition from each mouse; (D). Line plots presenting proportions of cell 928 proliferation activity per iron condition (percentage of total cells dividing 0, 1 929 or 2+ times) with log10 x-axes from each mouse; (E). Correlation plot 930 presenting the geometric means from each distribution versus the iron 931 condition for each mouse (with a log10 x-axis). Logarithmic trend lines, 932 GM = 762ag GM = 774ag GM = 775ag GM = 809ag GM = 853ag 0.625 mg/mL 0.125 mg/mL 0.025 mg/mL 0.005 mg/mL 0.001 mg/mL 0 1000 2000 0 30 60 90 120 0 50 100 0 50 100 0 20 40 60 0 20 40 60 Frequency 400 800 1200 0 20 40 60 80 100 400 800 1200 1600 0 20 40 60 80 100 Percentile 400 800 1200 1600 0 20 40 60 80 100 Mean Fe/cell (ag) 400 800 1200 1600 020406080100 Mean Fe/cell (ag) Holotransferrin (mg/mL) 0.625 0.125 0.025 0.005 0.001 Mouse 1 Mouse 2 Mouse 3 A A B C 0 20 40 60 0.0010.0100.1001.000 % of total cells Number of divisions 0 1 2+ 0.001 0.010 0.100 1.000 Holotransferrin (mg/mL) 0.001 0.010 0.100 1.000 0 20 40 60 0.001 0.010 0.100 1.000 % of total cells y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864 y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986 y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868 400 500 600 700 800 900 0.001 0.010 0.100 1.000 Holotransferrin (mg/mL) Fe mass/cell (ag) y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864y = 844 + 13 ⋅ l og10 x, r2 = 0.864 y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986y = 817 + 23 ⋅ l og10 x, r2 = 0.986 y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868y = 723 + 17 ⋅ l og10 x, r2 = 0.868 400 500 600 700 800 900 0.0010.0100.1001.000 Holotransferrin (mg/mL)Fe mass/cell (ag) Mouse 1 2 3 0e+00 1e+09 2e+09 3e+09 0.001 0.010 0.100 1.000 Holotransferrin (mg/mL) Total Fe (ag) r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 ppppppppppppppp = 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004 84 86 88 90 92 94 1.4 1.6 1.8 Skewness of Fe mass distributionCells divided (%) r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 r2 = 0.493 ppppppppppppppp= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004= 0.004 84 86 88 90 92 94 1.4 1.6 1.8 Skewness of Fe mass distribution Cells divided (%) Mouse 1 2 3 E F G GM = 652ag GM = 700ag GM = 729ag GM = 778ag GM = 799ag 0.625 mg/mL 0.125 mg/mL 0.025 mg/mL 0.005 mg/mL 0.001 mg/mL 0 1000 2000 0 20 40 60 80 0 20 40 60 80 0 25 50 75 0 30 60 90 0 10 20 30 Mass Fe (ag) GM = 627ag GM = 617ag GM = 643ag GM = 692ag GM = 725ag 0.625 mg/mL 0.125 mg/mL 0.025 mg/mL 0.005 mg/mL 0.001 mg/mL 0 1000 2000 0 25 50 75 100 125 0 30 60 90 120 0 50 100 150 200 0 25 50 75 0 50 100 0.001 - 0.625 mg/mL0.001 - 0.625 mg/mLHolotransferrin additionsHolotransferrin additions CD8+CD8+T-cellsT-cells SC-ICP-MSSC-ICP-MS FlowFlowcytometrycytometryNegativeNegativeisolationisolation OT-IOT-Imicemice IL-2, α-CD-3, α-CD28IL-2, α-CD-3, α-CD28 2 days2 days D preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint associated equations and r2 values are displayed for each profile. Error bars 933 show the 95th confidence intervals; (F). Line plot presenting the mean total 934 amount of iron consumed by the cells from the three mice for each iron 935 condition (with a log10 x-axis). Error bars show the range of total iron 936 calculated per mouse. (G). Scatter plot presenting the inverse correlation 937 between the magnitude of cell proliferation during the live culturing period 938 against the skewness of the single cell metallomic distributions for iron. 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 Fig. 4. Correlation of surface protein marker expression determined by Flow 961 Cytometry against SC-ICP-MS derived single cell metallomic contents. 962 (A). Mean fluorescence intensity (MFI) of CD71 (transferrin receptor) versus 963 the geometric mean single cell iron content (ag) for each population 964 distribution, from each holotransferrin condition; (B). Mean fluorescence 965 intensity (MFI) of CD25 (alpha chain component of the IL-2 surface receptor 966 on T-cells) versus the geometric mean single cell iron content (ag) for each 967 population distribution, from each holotransferrin condition; (C). Mean 968 fluorescence intensity (MFI) of CD71 versus the geometric mean single cell 969 calcium content (ag) for each population distribution, from each 970 holotransferrin condition; (D). Mean fluorescence intensity (MFI) of CD25 971 versus the geometric mean single cell calcium content (ag) for each population 972 distribution, from each holotransferrin condition. Error bars present the 95th 973 confidence intervals for each metallomic distribution determined by SC-ICP-974 MS. 975 976 1000 2000 3000 4000 500 700 900 Fe mass/cell (ag) CD71 MFI A 4000 6000 8000 500 700 900 Fe mass/cell (ag) CD25 MFI Mouse 1 2 3 B 1000 2000 3000 4000 0 10000 20000 Ca mass/cell (ag) CD71 MFI C 4000 6000 8000 0 10000 20000 Ca mass/cell (ag) CD25 MFI Mouse 1 2 3 D preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 Fig. 5. Results from primary human B-cell experiment. (A). An excerpt of time-999 resolved mass spectra measured for iron by PerkinElmer NexION 5000 SC-1000 ICP-MS from human B-cells taken from donor D1. Each spike represents a 1001 cell metallomic event, where their individual areas correlate to mass of iron 1002 within each cell. The metallomic plumes take ~0.001s to transit through the 1003 instrument (shown in the inset peak). The inset peaks also illustrate the 1004 resolutions achieved for the event profiles, where measurements were taken at 1005 intervals of 50µseconds. (B). Mass frequency histograms presenting 1006 distributions of single cell iron content determined by SC-ICP-MS for each of 1007 three donors, D1-D3, respectively. Mass distributions are shown in 1008 femtograms (fg) (10-15g). GM = distribution geometric mean for each 1009 population. (C). Histograms of CD19+ cell counts, determined by flow 1010 cytometry, for each of three donors, D1-D3, respectively. 1011 96.5% 95.8% 96.5% 93.6% 93.6% 93.6% 96.5% 95.1% Fluorescence minus one (FMO) Sample fluorescence A GM = 3.89fg GM = 3.25fg GM = 3.20fg D1 D2 D3 0 5 10 15 0 5 10 15 0 5 10 15 0 10 20 30 0 25 50 75 100 0 10 20 30 40 50 Mass Fe (fg) Frequency B C GM = 3.89fg GM = 3.25fg GM = 3.20fg D1 D2 D3 0 5 10 15 0 5 10 15 0 5 10 15 0 10 20 30 0 25 50 75 100 0 10 20 30 40 50 Mass Fe (fg) Frequency preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Table 1. Operating conditions for SC-ICP-MS and CyTOF. 1012 1013 1014 Parameter PerkinElmer NexION 350D ICP-MS PerkinElmer NexION 5000 ICP-MS Standard Biotools Helios CyTOF Plasma Gas (L/min) 18 16 17 Auxiliary Gas (L/min) 1.2 1.2 1.5 Makeup Gas (L/min) 0.80 0.60-0.75 0.65-0.95 Nebuliser Gas (L/min)) 0.25 0.25 0.15-0.17 RF Power (W) 1600 1600 1500 Sample Flow Rate (µL/min) 10 10 30 m/zAnalyte 103Rh 56Fe 40Ca 193Ir 103Rh 193Ir DRC Gas No gas NH3 NH3 No gas -/- -/- Flow Rate (mL/min) -/- 0.7 1.1 -/- -/- -/- RPq 0.25 0.7 0.3 0.25 -/- -/- Scanning mode MS MS/MS MS/MS MS/MS TOF TOF Dwell time (µs) Scan time (s) 75 60 50 180 40 60 10 60 13 3600 13 3600 Number of replicates 4 1 1 4 1 1 Injector Elemental Scientific 2mm quartz PerkinElmer 2mm quartz (White cassette) Lato Scientific WB injector Nebuliser Elemental Scientific CytoNeb 50 Meinhard HEN, Glass, 120psi Spray Chamber Elemental Scientific CytoSpray Standard Biotools Autosampler Elemental Scientific microFAST -/- preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Parameter PerkinElmer NexION 350D ICP-MS Plasma Gas (L/min) 18 Makeup Gas (L/min) -/- Auxiliary Gas (L/min) 1.2 Nebuliser Gas (L/min)) 0.915 RF Power (W) 1600 Sample Flow Rate (µL/min) ~250 m/zAnalyte 103Rh m/zInternal Standard 101Ru DRC Gas No gas DRC Gas Flow Rate (mL/min) -/- RPq 0.25 Mode MS Nebuliser Meinhard Plus Series quartz Injector 1.8mm quartz Spray Chamber Elemental Scientific cyclonic type Autosampler Elemental Scientific prepFAST M5 Table 2. Operating conditions for bulk-ICP-MS. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Supplementary Text

correction of SC-ICP-MS data Single cell inductively coupled plasma mass spectrometry (SC-ICP-MS) rapidly and precisely analyses individual cells from entire cell populations for their intracellular metal contents, thanks to its highly sensitive and fast measurement capabilities. However, SC-ICP- MS data is required to be collected successively, rather than averaged from numerous scans (as per bulk approaches), where corrections to backgrounds (which may exhibit fluctuations) is essential. Dissolved metals found within the extracellular media at levels above the limit of detection will be presented in the time-series data record as a constant signal that is elevated above baseline (25). If this signal is stable, systematic 3σ + mean background corrections maybe all that is required to reduce the data to exclusively reveal only the single cell results - as conducted in the Single Cell module of the PerkinElmer Syngistix software. However, other irregular background entities found within the SC-ICP-MS data are much trickier to resolve. An example includes peaks in the data recording fragments of lysed cells and/or those cells that are found below detection. In our study we present the ability to discriminate between such entities and whole cells in the SC-ICP-MS data by virtue of their transit time through the mass spectrometer. Through our rapid scanning procedure, where measurement frequencies are conducted at time-scales that are significantly shorter than the cell events, we are distinctly able to resolve differences between metallomic event durations (and thus whole cells versus cell fragments/ cells below detection). In the experiments that we conducted in this study, our threshold was selected by the lowest peak amplitude that was able to sustain a whole cell event. We found that whole cells above the threshold were presented with well- constrained cell event duration times of ~0.5ms - when no cell gas is used, to ~1ms - when preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint the cell is pressurised with NH3. This contrasts those events below the threshold (containing events from cell fragments from lysed cells and/or cell events below detection), which instead transited through the mass spectrometer at much faster time-scales. An example is presented in Fig. S1, where the “non-cell” event in these measurements for iron exhibits a duration time that is more than 30% shorter than the example “cell event”. Verification towards the accuracy of the background corrections that we applied are highlighted from the transport efficiency data reported from the murine T-cell experiment, where distinct similarities in such values between those collected from calcium to those from iron were made (overall averages being 12.2% versus 12.7% respectively) (see Table S2). Moreover, as these values are also similar to those transport efficiencies reported from related studies (e.g. 15, 26, 27), this provides further validation of the method of background correction that we applied. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Figs. S1 – S2 Fig. S1. Time-resolved single cell mass spectra measured by SC-ICP-MS from murine T-cells. Measurement of iron in T-cells taken from mouse 3 and the 0.005mg/mL condition. Each spike represents an individual metallomic event, where their individual areas correlate to the mass of iron. The inset taken at ~0.76 seconds highlights a cell event, where the metallomic plume transits through the mass spectrometer within ~1ms. The inset taken at ~7.18 seconds highlights a fragment from a lysed cell/ cell event below detection. Here the metallomic plume transits through the instrument over a much shorter time- span (~0.65ms), thus rendering it below the detection threshold. Fig. S2. Correlation of inter-quartile ranges against holotransferrin iron-media condition in the murine T-cell experiment. Logarithmic trend lines, associated equations and r2 values are also displayed. y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529y = 439 + 9.3 ⋅ l og10 x, r2 = 0.529 y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9y = 359 + 9.3 ⋅ l og10 x, r2 = 0.9 y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8y = 321 + 12 ⋅ l og10 x, r2 = 0.8 0 100 200 300 400 500 0.001 0.010 0.100 1.000 Holotransferrin (mg/mL) Interquartile range (ag) Mouse 1 2 3 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Tables S1 – S3 Intercalation molarity (μmol) Metal intercalator Interquartile range (ag) 0.005 Iridium 21 0.01 7.5 0.02 15 0.03 24 0.04 27 0.05 36 0.05 Rhodium 50 0.1 48 0.2 104 0.3 225 0.4 320 0.5 385 Table. S1. Correlation of inter-quartile ranges against intercalation molarity from the metal-intercalation experiments. Mouse Holotransferrin condition (mg/mL) Transport efficiency: T-cell Fe-derived (%) Transport efficiency: T-cell Ca- derived (%) Transport efficiency: EQTM Four-Element Calibration Beads: Ho-derived (%) n/a n/a -/- -/- 29.1 n/a n/a -/- -/- 22.9 1 0.001 9.6 10.8 -/- 0.005 7.8 11.8 -/- 0.025 10.3 8.8 -/- 0.125 16.3 8.8 -/- 0.625 19.0 15.7 -/- 2 0.001 8.1 13.2 -/- 0.005 21.6 18.2 -/- 0.025 10.4 10.6 -/- 0.125 17.0 17.2 -/- 0.625 12.6 15.5 -/- n/a n/a -/- -/- 25.9 3 0.001 10.8 10.9 -/- 0.005 11.1 7.3 -/- 0.025 15.7 14.5 -/- 0.125 7.9 9.7 -/- 0.625 11.5 10.4 -/- n/a n/a -/- -/- 28.9 Average 12.7 12.2 26.7 Table S2. Transport efficiency data from the murine T-cell experiments. All transport efficiency data calculated by the particle frequency method, as described in (25). Both iron and calcium-derived transport efficiencies were derived directly from the murine T-cells, and holmium-derived transport efficiencies from EQTM Four-Element Calibration Beads. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint Mouse Holotransferrin condition (mg/mL) Live cell count A (cells/mL) Live cell count B (cells/mL) Mean (A and B) (cells/mL) Total volume (mL) Total number of cells 1 0.001 1.99E+05 2.77E+05 2.38E+05 7.5 1.79E+06 0.005 2.67E+05 3.40E+05 3.04E+05 7.5 2.28E+06 0.025 3.56E+05 2.72E+05 3.14E+05 7.5 2.36E+06 0.125 2.04E+05 2.25E+05 2.15E+05 7.5 1.61E+06 0.625 2.30E+05 2.83E+05 2.57E+05 7.5 1.92E+06 2 0.001 1.88E+05 2.77E+05 2.33E+05 6 1.40E+06 0.005 3.61E+05 3.98E+05 3.80E+05 6 2.28E+06 0.025 3.35E+05 3.03E+05 3.19E+05 6 1.91E+06 0.125 2.20E+05 2.67E+05 2.44E+05 6 1.46E+06 0.625 2.93E+05 2.56E+05 2.75E+05 7.5 2.06E+06 3 0.001 4.92E+05 5.28E+05 5.10E+05 7.5 3.83E+06 0.005 5.91E+05 5.02E+05 5.47E+05 7.5 4.10E+06 0.025 4.39E+05 5.81E+05 5.10E+05 7.5 3.83E+06 0.125 4.81E+05 4.50E+05 4.66E+05 7.5 3.49E+06 0.625 3.92E+05 4.08E+05 4.00E+05 7.5 3.00E+06 Table S3. Live cell counts and total number of cells harvested following the 2-day culturing period for the murine T-cell experiment. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted November 14, 2024. ; https://doi.org/10.1101/2024.11.11.623006doi: bioRxiv preprint