Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Jun 1.
Published in final edited form as: Magn Reson Med. 2019 Nov 25;83(6):2276–2283. doi: 10.1002/mrm.28081

Effect of Mouse Strain and Diet on Feasibility of MRI-Based Cell Tracking in the Liver

Christiane L Mallett 1, Jeremy ML Hix 1, Matti Kiupel 2, Erik M Shapiro 1
PMCID: PMC7047545  NIHMSID: NIHMS1057029  PMID: 31765493

Abstract

Purpose:

MRI-based cell tracking identifies the location of magnetically labeled cells with hypointense voxels. Here we demonstrate a strain-dependent effect of liver MRI background on the feasibility of MRI-based cell tracking of transplanted cells in the mouse liver.

Methods:

FVB mice (GFP-LUC and NOG) and C57Bl/6 mice (GFP+ and wild-type) were fed three different diets with varying iron content. In vivo T2*-weighted images and T2* maps of the liver were acquired at different ages. Magnetically labeled cancer cells were injected intrasplenically for hepatic migration, then mice were imaged by in vivo MRI and BLI. Livers were also imaged ex vivo by MPI.

Results:

R2* increased with age in FVBNOG and FVBGFP-LUC mice fed diets sufficient in iron. FVBNOG mice developed a mottled appearance in their livers with age that did not occur in FVBGFP-LUC mice. R2* was unchanging with age in C57Bl/6GFP mice and the liver remained bright and homogenous. Labeled cells were not detectable by MRI in some livers despite successful engraftment as shown by BLI and MPI.

Conclusions:

Strain, diet and age are important considerations for MRI-based cell tracking in the liver. If a model with excessive liver iron must be used, alternative imaging methods such as MPI can be considered.

Keywords: Liver, Iron, ferritin, cell tracking, mouse, MPIO

INTRODUCTION

MRI-based cell tracking is a sensitive, non-invasive method to detect cells labeled with iron oxide nanoparticles. In this method, cells are labeled in vitro or in vivo with superparamagnetic iron oxide particles. The resulting local magnetic field inhomogeneity accelerates spin dephasing, resulting in hypointense signal on gradient-echo images and enabling an ultimate cellular detection sensitivity of single cells (1,2). MRI-based cell tracking generally performs better in tissues with homogenous imaging features such as the liver, which has been used as a model tissue for single-cell detection of transplanted hepatocytes (1), pancreatic islets (3,4) and mesenchymal stem cells (5,6).

Here we describe a phenotype of hepatic iron accumulation in some strains of mice that makes MRI-based cell tracking challenging. Immunocompromised mice such as NOG (NOD/SCID IL-2 γ receptor knockout), which have no T-cells, B-cells or natural killer cells, can be used as recipients of xenotransplants. We observed a pattern of dark and mottled MRI signal in livers in NOG mice and FVB mice, especially with increased age, which did not occur with mice on a C57Bl/6 background. We used T2*-weighted imaging and T2* maps to measure the increase in hepatic iron with age in mice from FVB and C57Bl/6 backgrounds that were fed a variety of diets. Then, we transplanted mice with iron oxide labeled cells to evaluate the effect of the increased hepatic iron on the ability to detect labeled cells. We then used magnetic particle imaging (MPI) to detect the labeled cells ex vivo.

METHODS

Mice

Animal experiments were approved by our IACUC under protocols 2015-08-132-00 and 201800027. Mice were housed a specific-pathogen free facility (C57Bl/6WT were conventionally-housed) in an AAALAC-accredited vivarium, with 12 hour light/dark cycles and food and water ad libitum. We studied NOG mice, with GFP-Luciferase mice also from an FVB background as controls, and compared them to GFP mice with a C57Bl/6 background and wild-type C57Bl/6 mice. The control strains were chosen based on availability of colonies at our institution. Mice were fed a standard higher iron diet, a standard medium iron diet or a specially-ordered low iron diet (Table 1).

Table 1: Summary of mouse strains and diets.

NOG mice were from a colony of NOG/TK-NOG mice under license (Taconic Biosciences model 12907, Hudson, USA). TK-NOG mice express herpes simplex virus type 1 thymidine kinase so that administration of the pro-drug ganciclovir specifically ablates hepatocytes (18). We did not administer ganciclovir to any mice in this study, so NOG and TK-NOG mice are pooled together. The TK-NOG mice were heterozygous for the TK gene. NOG mice are bred on an FVB background, so an in-house colony of mice on that background were used as an immune-competent comparison (Jackson labs, FVB-Tg(CAG-luc,-GFP)L2G85Chco/J stock number 008450). A colony of mice on a C57Bl/6 background with ubiquitous GFP expression (Jackson labs, C57bl/6-TG(UBC-GFP)30Scha/J stock number 004353) and older C57Bl/6 mice (Charles River) were also examined. All diets are from Envigo. High iron diet is either 7913 or 8940 (270–280 mg Fe/kg). Medium iron is 2919 (200 mg Fe/kg). Low iron diet is TD.99397 (8 mg Fe/kg).

Name Strain Background Iron in diet (mg/kg) Study Type Age (wk) N (M)
FVBNOG-hiFe (TK)-NOG FVB 270/280 Cross-sectional 9–31 27 (11)
FVBNOG-medFe (TK)-NOG FVB 200 Cross-sectional 7–40 38 (22)
FVBNOG-medFe (TK)-NOG FVB 200 Detection of cancer cells 7–32 18 (10)
FVBNOG-lowFeYoung (TK)-NOG FVB 8 (at weaning) Longitudinal 8–24 12 (6)
FVBNOG-lowFeOld (TK)-NOG FVB 8 (as adults) Longitudinal 20–31 7 (0)
C57Bl/6GFP-medFe C57Bl/6-GFP C57Bl/6 200 Cross-sectional 9–64 21 (9)
FVBGFP-LUC-hiFe FVB-GFP-luc FVB 270/280 Cross-sectional 7–25 17 (8)
C57Bl/6WT-hiFe C57BL/6 C57Bl/6 270/280 Cross-sectional 51 4 (4)
Open in a new tab

Magnetic resonance imaging

Images were acquired on a 7T Biospec 70/30 USR (Bruker) using a 9 cm diameter volume transmit coil and four channel surface array receive coil (4×4 cm (diet studies) or 2×2 cm (cell tracking study)), with respiratory gating. Mice were anesthetized using 1.5–3% isoflurane in oxygen. Temperature (~35 °C) and breathing (30–60 bpm) were monitored throughout (SAII). T2*-weighted FLASH images parameters: TR/TE=300/6 ms, flip angle 30°, 10 averages, 100×100×300 μm. T2* map (multi-gradient echo) parameters: TR/TE=1500/2.02 ms, 8–12 positive echoes acquired every 1.63 ms, 200×200×300 µm. Total acquisition time was 1h. T2* maps were analyzed in the image sequence analysis module of Paravision 6.0.1; for low T2* values, fewer points were fit to ensure a better exponential fit (5–10 points instead of 12). We measured T2* in three regions of interest in different slices, then calculated the mean T2*. R2* was then calculated as 1/T2*.

Ex-vivo analysis

We euthanized mice by CO2 overdose followed by cervical dislocation within 1 week after their last scan. Immediately after euthanasia, we collected cardiac blood into a capillary tube, spun the tubes at 10,000g for 5 minutes, then measured the height of the red blood cell fraction (RBC) compared to the total blood height and calculated hematocrit as RBC/total*100. Livers were fixed in 10% formalin for 24 h then 30–70% ethanol for paraffin embedding and sectioning. After staining with H&E and Perls’ Prussian blue, sections were examined by a veterinary pathologist. Bright field images were acquired at 5x on a DMI 4000B microscope (Leica). After an unblinded survey of images, 5 patterns of iron deposition were identified and assigned scores: no iron (0), light spotting throughout (1), light deposition between lobes (2), heavy deposition throughout (3), heavy deposition between lobes (4). Images from three sections each from 15 FVBNOG-hiFe mice and 11 FVBGPF-LUC-hiFe mice were then anonymized in FIJI (7) using the Blind Analysis macro (from http://quantixed.org). The pattern of iron deposition in the blinded images was scored by one observer three times. The mode of the three scores was used as the final score for each section.

MRI-based cell tracking

To investigate the effect of liver background MRI appearance on ability to detect iron-labeled cells, we transplanted magnetically labeled MDA-MB-231-luc cells (a kind gift from Dr. Anna Moore) into FVBNOG-medFe mice. Cells were grown in DMEM (Gibco, Fisher) with 10% fetal bovine serum at 5% CO2. They were plated in T150 flasks and labeled overnight with 1.63 µm Bangs particles (Bangs Laboratories) at 70% confluence. Cells were harvested by trypsinization, washed 3x in phosphate-buffered saline (PBS) then suspended at 20×106 cells/mL. For the surgery, mice were anesthetized with 2% isoflurane, the spleen was exposed and partially exteriorized, and 1×106 cells in 50 µL of PBS were injected into the spleen. The spleen was then re-inserted and the incision was sutured. Buprenorphine and meloxicam were injected for analgesia. We acquired in vivo bioluminescence imaging (BLI) and liver MRI on day 3–4 and 8–9 post-surgery. BLI (IVIS Spectrum, Perkin-Elmer) images were acquired 10 minutes after subcutaneous injection of luciferase (Gold Biosciences, St Louis MO USA) at 150 mg/kg. After euthanasia, excised livers were scanned in a Momentum magnetic particle imager (Magnetic Insight) using the 2D isotropic setting.

Statistics

We used GraphPad Prism 7.04 for all statistical analysis (Graph Pad Software). For R2* vs age we calculated linear regression, then used the calculated slopes and standard error of the mean in a one-way analysis of variance and a Tukey-Kramer post-hoc test to compare the slopes to each other. To determine the effect of a low-iron diet, we used a one-way analysis of variance with Dunnet’s post-hoc test to compare each time point to the first one. We did not use a repeated measures analysis of variation because not all mice were imaged at all timepoints. We used a one-way analysis of variance to compare hematocrit between strains and diets in each age group, followed by a Tukey-Kramer post-hoc test. We did not use a two-way analysis of variance because not all strains had hematocrit measured at all ages. The significance threshold was set as p<0.05 for all assays.

RESULTS

T2*-weighted imaging

The MRI appearances of the livers of all mice were similar when they were young but diverged with age (Figure 1). The MRI of livers in mice on an FVB background darkened with age; FVBNOG mice developed a mottled appearance in their livers, while the FVBGFP-LUC-hiFe mice had a uniform liver appearance. Feeding FVBNOG mice a low-iron diet from weaning prevented dark spots in the MRI from developing, while feeding older mice a low-iron diet slightly decreased the appearance of the spots but they persisted. Mice on a low-iron diet from birth had a reduced hematocrit (Supporting Information Figure S1) and needed to be euthanized at age 16–20 weeks because of poor body condition. The liver appearance of the C57Bl/6GFP-medFe mice were unchanged as they aged, and were similar for aged C57bl/6WT-hiFe mice. Young (~6–12 week) FVBNOG-medFe and FVBGFP-LUC-hiFe had some hypointensities in their liver MRI, especially the lower lobes, which disappeared as they aged. When FVBNOG mice were fed a medium iron diet, there was still a mottled pattern in their livers, but not as severe as when they were fed a higher iron diet (see Figure 4).

Figure 1: T2*-weighted images of mouse livers in young and old mice.

Figure 1:

Open in a new tab

Younger mice are in the left column, older mice in the right column. a) 12 week old FVBNOG-hiFe, b) 22 week old FVBNOG-hiFe, c) 11 week old FVBGFP-LUC-hiFe, d) 22 week old FVBGFP-LUC-hiFe, e) 13 week C57Bl/6GFP-medFe, f) 47 week old C57Bl6GFP-medF, g) 18 week old FVBNOG-lowFeYoung, h) 29 week old FVBNOG-lowFeOld, i) 9 week old FVBGFP-LUC-hiFe showing hypointensities in lower lobes, j) 51 week old C57Bl/6WT-hiFe. Scale bar is 0.5 cm.

Figure 4: Ability to detect cell transplants by MRI is compromised in some FVBNOG livers.

Figure 4:

Open in a new tab

Images are from two representative mice fed a medium iron diet. Top row is a male mouse age 7 wk, bottom row is from a female mouse age 30 wk. a, e) T2*-weighted image at baseline. b,f) T2*-weighted image 4 days after transplant of cells labeled with MPIO. Hypointensities indicate the presence of the iron labeled cells. c, g) BLI from 9 days after transplant showing the proliferation of the labeled cancer cells. The spleen (injection site) and liver have live cells in them. d, h) MPI images of liver only (ex vivo). Scale bars are 0.5 cm (for MRI and MPI).

Relaxometry

R2* increased with age for all mice on an FVB background, regardless of diet (Figure 2a and 2b), from approximately 100 s−1 at age 10 weeks to 200–300 s−1 at age 35 weeks. 7 male FVBNOG-medFe and 1 female FVBNOG-hiFe aged 7–9 weeks were excluded from the R2* analysis because of the widespread hypointensities in their livers at a young age (mean R2* 150 s−1, range 110–250 s−1). C57bl/6GFP-medFe mice had less of an increase in R2* with age; it ranged from approximately 65 s−1 at age 10 weeks to 75–120 s−1 at age 60 weeks. For all strains, the correlation between age and R2* was better when male and female mice were analysed separately, with the female mice generally having higher R2* values (Supporting Information Table S1a). When the male and female mice were pooled together, the FVBGFP-LUC-hiFe and FVBNOG-hiFe slopes were significantly higher than the C57Bl/6GFP-medFe slope. When the male and female mice were separated, the significant differences were: FVBGFP-LUC-hiFe-F > C57Bl/6GFP-medFe-F and –M; FVBNOG-hiFe-F>FVBGFP-LUC-hiFe-M, FVBNOG-medFe-F and FVBNOG-medFe-M, FVBNOG-hiFe-M, C57Bl/6GFP-medFe-F, C57Bl/6GFP-medFe-M (Table S1b).

Figure 2: Relationship between age and liver R2* by strain, sex and diet.

Figure 2:

Open in a new tab

a) R2* vs age for cross-sectional studies for male mice. b) R2* vs age for cross-sectional studies for female mice. Lines show linear regression. c) Liver R2* of FVBNOG mice fed a low iron diet from 6 weeks old. d) Liver R2* of FVBNOG mice fed a low iron diet from 20 weeks old. For c and d, each point is a single mouse and the lines show the mean ± standard deviation. * indicates a significant difference (p<0.05) from the first measurement.

In FVBNOG mice fed a low iron diet from 6 weeks of age, liver R2* decreased from approximately 100 s−1 at 8 weeks old to about 60 s−1 at 26 weeks (Figure 2c). It was significantly lower than baseline values from 12 weeks of age onwards. When FVBNOG mice were switched from a high iron diet to a low iron diet at 20 weeks old, liver R2* was consistent over time (Figure 2d). Mean R2* at age 20 weeks was 155 s−1 and there was only a significant decrease from that at age 27 weeks (118 s−1).

Histology

After a blinded scoring of mice on a high iron diet from an FVB background, all sections that were scored with heavy deposition throughout (score 3) were from FVBGFP-LUC-hiFe mice, while all sections that had heavy staining along the edges of the lobes were from FVBNOG-hiFe mice (score 4) (Figure 3). FVBNOG-medFe mice had similar heavy deposition of iron on the edges of the lobes, while C57bl/6GPF mice had minimal iron staining, and it was usually around the big vessels. There were no pathologically significant structural abnormalities and the amount of hepatic iron was not pathologically significant in any mouse. In FVBNOG-hiFe mice, there were incidental findings of mild centrolobular vacuolorization, pigment granulomas (which sometimes coincided with iron deposition), extramedullary hematopoiesis, and lipofusin. In FVBGFP-LUC-hiFe mice, there were incidental findings of perivascular inflammation and necrosis and idiopathic necrosis. In C57bl/6GFP-medFe mice, there were pigment granulomas and a few apoptotic hepatocytes. (Supporting Information Figure S2).

Figure 3: Difference in distribution of iron by strain.

Figure 3:

Open in a new tab

a) Percentage of sections with each iron score, by strain. 39 sections from 11 FVB-GFP-LUC-hiFe mice and 53 sections from 15 FVBNOG-hiFe mice were scored. b) Representative Perls’ stained section that was scored 3 (large spots) from a female 35 week old FVBGFP-LUC-hiFe mouse. c) Representative Perls’ Prussian Blue-stained section that was scored 4 (heavy deposition along lobes) from a 39 week old female FVBNOG-hiFe mouse. Scale bar is 200 µm.

MRI-based cell tracking

We transplanted iron oxide labeled cancer cells into the spleen to demonstrate the effect of liver appearance on the ability to detect iron labeled cells by MRI (Figure 4). Cells transplanted into the spleen will seed the liver via the vasculature (1). In a representative young male mouse with no spots in the baseline liver MRI, the transplanted cells are clearly visible as hypointense spots, while they are not visible at all in an older female mouse with a darker liver pre-transplant. Both images are from FVBNOG-medFe mice. We confirmed that there was successful migration to the liver and engraftment using BLI. When the livers were excised and scanned ex vivo using MPI, the iron labeled cells were detected in both mice. There was no background MPI signal in an untransplanted mouse (Supporting Information Figure S3).

DISCUSSION AND CONCLUSIONS

MRI-based cell tracking has been used to track a variety of transplanted cells including hepatocytes, cancer cells, stem cells and infiltrating macrophages (1,2,8,9). Generally, a desirable tissue for MRI-based cell tracking is relatively bright, with uniform background such as brain, cartilage or liver. We describe MRI liver appearance in some strains of mice that could make MRI-based cell tracking in the liver challenging in young and old mice.

We acquired T2* weighted FLASH images and T2* maps of mice from three strains and diets: FVBNOG mice fed a conventional higher iron diet and a conventional medium iron diet, FVBGFP-LUC mice fed a conventional higher iron diet, C57bl/6GFP mice fed a medium iron diet, and a few C57bl/6WT mice fed a higher iron diet. In younger FVBNOG-hiFe and -medFe and FVBGFP-LUC-hi-Fe mice, there were hypointensities from iron deposition in the lower lobes (which would interfere with MRI-based cell tracking); they resolved as the mice aged to about 12 weeks. As the mice aged, livers became more hypointense, and in older FVBNOG-hiFe and -medFe mice, there was a mottling pattern, which would also interfere with MRI-based cell tracking.

R2* increased linearly with age in the FVBNOG and FVBGFP-LUC mice, while remaining low in the C57bl/6GFP mice. R2* is linearly related to liver iron (10). Feeding FVBNOG mice a very low iron diet resolved the hypointensities, but this diet was too low in iron for normal husbandry. This very low iron diet was chosen to confirm that the MRI phenotype in FVBNOG mice was related to liver iron and to attempt to resolve the spottiness in the older FVBNOG mice. In the older mice fed a low-iron diet, there was not a significant decrease in R2*, perhaps because there was sufficient iron stores in the body and there was no need to draw down from the liver stores. It is possible that a diet that is lower in available iron than the two conventional diets (but still providing sufficient dietary iron) would sufficiently remove the hypointensities and allow for cellular imaging at all ages. The recommended amount of iron in a mouse diet is ~50 mg/kg (11). The medium and high iron diets used in this study were all conventional plant-based diets fed to mice at our institution and thus should be adequate for animal health. It is difficult to predict bioavailable iron from iron content in these diets; dietary iron concentration is not the final determinant of iron bioavailability. For example, a plant-based diet comparable to the ones used in our study (high iron content but low bioavailability) had a total iron content of 200 mg/kg but resulted in similar iron load to mice fed an iron replete diet with high bioavailability (45 mg/kg) (12). However, in female mice, there was more iron accumulation in the FVBNOG-hiFe livers than the FVBNOG-medFe livers, so the difference in bioavailable iron between diets may be significant. The three diets had similar calcium (0.9–1.1%) and phytates (0.3%) and total fiber (14.7–16.7%) (13–15).

Microscopically, the iron deposition pattern was different between FVBGFP-LUC-hiFe mice, which had iron deposits throughout the liver, and FVBNOG-hiFe mice, which had perilobular iron. This iron deposition pattern may account for the mottling effect seen in MRI of the FVBNOG-hiFe mice. Importantly, none of the iron loading was deemed to be clinically significant. Iron is stored in the liver in ferritin and hemosiderin, both of which have superparamagnetic cores (16,17) and are therefore detectable by MRI.

There was more iron in females than males, and more accumulation of iron in FVB mice than C57Bl/6 mice. This is consistent with previous studies which have found strain-, sex- and age-dependent differences in mouse liver iron. Little to no work has been done with FVB mice, but C57Bl/6 are consistently at the lower end of liver iron relative to other strains. Liver iron was elevated in aged C57Bl/6 and BALB/c mice compared to young adult mice (18). In BALB/c and DBA/2J mice, females had more liver iron than males, while there was no difference between the sexes in C57Bl/6 mice (18). The authors suggest the difference between males and females may be regulated by hormones. Another study found higher liver iron in female C57Bl/6 mice vs males at age 14 weeks, while their plasma iron was the same between sexes (19). C57Bl/6 mice cleared iron from the circulation into the liver more slowly than SWR mice (20). There are also sex and/or strain differences in liver expression of mRNA of Hamp1, although the relationship between plasma hepcidin, Hamp1 expression and liver iron is still uncertain (21,22).

Labeled cells were detectable by MRI in young mice with bright livers pre-transplant, but not in mice with darker livers, even when BLI and MPI confirmed that labeled cells were present. This suggests that MPI could be used as an alternative to MRI to detect the presence of iron labeled cells in cases where there is a confounding background signal. We used ex vivo MPI, but in vivo scans at the same settings are routine. MPI detects only superparamagnetic iron with no anatomy as background, and images have positive contrast, unlike the negative contrast obtained in MRI; the spatial resolution (~1 mm) is currently coarser than that of MRI (23). Another quantifiable, positive contrast option would be to use a fluorine-based label and 19F MRI (24,25), which has a similar spatial resolution to MPI. One advantage that both MPI and 19F MRI have over proton MRI is the positive contrast and dynamic range – it is hard to differentiate between high amounts of iron in dark proton MRI images.

MRI-based cell tracking has been performed in the liver in C57Bl/6 mice (hepatocytes) (1) and in nude mice (islet transplant in streptozotocin induced diabetes) (4). In those studies, there was no indication of excess iron accumulation interfering with detection of the labeled cells, which is consistent with our observations in C57Bl/6 mice.Islet transplants have also been detected in NOD.SCID mice, which are related to the FVBNOG mice used in this study (3). In that study, the mice were followed for 2 weeks post-transplant, and their age and diet was not specified. Thus, depending on the experimental design and application, it is possible to work around the iron deposition pattern we have identified. This effect may not be important in experiments in mice that are between 12–16 weeks, but we have demonstrated that in young and old mice, there are naturally-occurring iron deposits that could be significant confounds, especially for long-term MRI-based cell tracking.

CONCLUSIONS

We have identified a strain-dependent difference in iron deposition patterns in mouse liver. This has important ramifications for MRI-based cell tracking, especially for long-term detection of transplanted cells. This finding demonstrates the need to select models and appropriate controls carefully, and consider factors such as diet in experimental design. Alternative imaging strategies such as magnetic particle imaging could be considered for robust long term detection of magnetically labeled cells in vivo.

Supplementary Material

Supp info

Supporting Information Table S1a: Correlation coefficients and slopes for R2* vs age, with mice divided by sex and pooled together.

Supporting Information Table S1b. Summary of comparison between slopes of R2* vs age for mice of different strains and diets. A one-way analysis of variance was performed and was significant (p<0.05), then a Tukey-Kramer post-hoc test was used to compare each slope to the others.

Supporting Information Figure S1: Hematocrit for mice from FVB background. Hematocrit was measured for a subset of mice at euthanasia. Mice were grouped by age, strain, and diet. Differences in hematocrit were measured by analysis of variance within age groups, not between ages. The analysis of variance was significant for both analyses (p<0.05). Letters indicate strains that are statistically the same after Tukey’s post-hoc test for multiple comparisons (p<0.05). Dotted line indicates mean hematocrit of 46.6% for 16 week old mice (JAX data sheet-http://jackson.jax.org/rs/444-BUH-304/images/physiological_data_001800.pdf). Numbers at base of bars indicate the number of mice in each group. The hematocrit for the FVBNOG mice on the low-iron diet was significantly lower than the FVBNOG mice on the high iron diet. Mice that were found dead in the cage (presumably due to complications from anemia) or were euthanized before 10 weeks are not included in this measurement so this may be an underestimate of the true effect of the low iron diet on hematocrit.

Supporting Information Figure S2: Additional Histology a) H&E stained section from a 35 wk old FVBGFP-LUC-hiFe mouse with extramedullary hematopoiesis (arrow head). b) An adjacent PPB-stained section where there is no iron deposition corresponding to the EMH. c) H&E stained section from 39 wk old FVBNOG-hiFe mouse with pigment granulomas (arrows). d) An adjacent PPB-stained section where some of the iron staining is co-localized with the pigment granulomas. E) An example of a section that scored a 1 (light spotting throughout) on the PPB scoring system (from a 39 wk old FVBNOG-hiFe mouse). f) An example of a section that scored a 2 (light deposition between lobes) on the PPB scoring system (from a 38 wk old FVBNOG-hiFe mouse). Scale bar is 200 μm.

ACKNOWLEDGEMENTS

The authors thank Dr. John Zubek for assistance with measuring hematocrit, Dr. Anna Moore for the MDA-MB-231-luc cells, and funding from NIH R01 DK107697.

REFERENCES

  • 1.Shapiro EM, Sharer K, Skrtic S, Koretsky AP. In vivo detection of single cells by MRI. Magn Reson Med. 2006;55(2):242–9. [DOI] [PubMed] [Google Scholar]
  • 2.Heyn C, Ronald JA, Mackenzie LT, MacDonald IC, Chambers AF, Rutt BK, et al. In vivo magnetic resonance imaging of single cells in mouse brain with optical validation. Magn Reson Med. 2006. January 1;55(1):23–9. [DOI] [PubMed] [Google Scholar]
  • 3.Evgenov NV, Medarova Z, Pratt J, Pantazopoulos P, Leyting S, Bonner-Weir S, et al. In Vivo Imaging of Immune Rejection in Transplanted Pancreatic Islets. Diabetes. 2006. September 1;55(9):2419–28. [DOI] [PubMed] [Google Scholar]
  • 4.Evgenov NV, Medarova Z, Dai G, Bonner-Weir S, Moore A. In vivo imaging of islet transplantation. Nat Med. 2006. January;12(1):144–8. [DOI] [PubMed] [Google Scholar]
  • 5.Lai L, Chen J, Wei X, Huang M, Hu X, Yang R, et al. Transplantation of MSCs Overexpressing HGF into a Rat Model of Liver Fibrosis. Mol Imaging Biol. 2016. February 1;18(1):43–51. [DOI] [PubMed] [Google Scholar]
  • 6.Choi D, Kim JH, Lim M, Song KW, Paik SS, Kim SJ, et al. Hepatocyte-Like Cells from Human Mesenchymal Stem Cells Engrafted in Regenerating Rat Liver Tracked with In Vivo Magnetic Resonance Imaging. Tissue Eng Part C Methods. 2008. March;14(1):15–23. [DOI] [PubMed] [Google Scholar]
  • 7.Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, et al. Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012. July;9(7):676–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Daldrup-Link HE, Golovko D, Ruffell B, Denardo DG, Castaneda R, Ansari C, et al. MRI of tumor-associated macrophages with clinically applicable iron oxide nanoparticles. Clin Cancer Res Off J Am Assoc Cancer Res. 2011. September 1;17(17):5695–704. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Barczewska M, Wojtkiewicz J, Habich A, Janowski M, Adamiak Z, Holak P, et al. MR monitoring of minimally invasive delivery of mesenchymal stem cells into the porcine intervertebral disc. PloS One. 2013;8(9):e74658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Wood JC, Enriquez C, Ghugre N, Tyzka JM, Carson S, Nelson MD, et al. MRI R2 and R2* mapping accurately estimates hepatic iron concentration in transfusion-dependent thalassemia and sickle cell disease patients. Blood. 2005. August 15;106(4):1460–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Kautz L, Gabayan V, Wang X, Wu J, Onwuzurike J, Jung G, et al. Testing the Iron Hypothesis in a Mouse Model of Atherosclerosis. Cell Rep. 2013. December;5(5):1436–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hubbard AC, Bandyopadhyay S, Wojczyk BS, Spitalnik SL, Hod EA, Prestia KA. Effect of Dietary Iron on Fetal Growth in Pregnant Mice. Comp Med. 2013. April;63(2):127–35. [PMC free article] [PubMed] [Google Scholar]
  • 13.Teklad Global 19% Protein Extruded Rodent Diet. Envigo; website. https://www.envigo.com/resources/data-sheets/2019-datasheet-0915.pdf. Updated 2015, accessed October 23, 2019. [Google Scholar]
  • 14.Teklad 22/5 Rodent Diet. Envigo; website. https://www.envigo.com/resources/data-sheets/8640-datasheet-0915.pdf Updated 2015, accessed October 23, 2019. [Google Scholar]
  • 15.NIH-31 Modified Open Forula Mouse/Rat Sterilizable Diet. Envigo; website. https://www.envigo.com/resources/data-sheets/7013-datasheet-0915.pdf. Updated 2015, accessed October 23, 2019. [Google Scholar]
  • 16.Owen CS, Lindsay JG. Ferritin as a label for high-gradient magnetic separation. Biophys J. 1983. May 1;42(2):145–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Gottesfeld Z, Neeman M. Ferritin effect on the transverse relaxation of water: NMR microscopy at 9.4 T. Magn Reson Med. 1996;35(4):514–20. [DOI] [PubMed] [Google Scholar]
  • 18.Hahn P, Song Y, Ying G, He X, Beard J, Dunaief JL. Age-dependent and gender-specific changes in mouse tissue iron by strain. Exp Gerontol. 2009. September;44(9):594–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Unger EL, Beard JL, Jones BC. Iron regulation in C57BLI6 and DBA/2 mice subjected to iron overload. Nutr Neurosci. 2007. April;10(1–2):89–95. [DOI] [PubMed] [Google Scholar]
  • 20.Clothier B, Robinson S, Akhtar RA, Francis JE, Peters TJ, Raja K, et al. Genetic variation of basal iron status, ferritin and iron regulatory protein in mice: potential for modulation of oxidative stress. Biochem Pharmacol. 2000. January 15;59(2):115–22. [DOI] [PubMed] [Google Scholar]
  • 21.Courselaud B, Troadec M-B, Fruchon S, Ilyin G, Borot N, Leroyer P, et al. Strain and gender modulate hepatic hepcidin 1 and 2 mRNA expression in mice. Blood Cells Mol Dis. 2004. March 1;32(2):283–9. [DOI] [PubMed] [Google Scholar]
  • 22.McLachlan S, Page KE, Lee S-M, Loguinov A, Valore E, Hui ST, et al. Hamp1 mRNA and plasma hepcidin levels are influenced by sex and strain but do not predict tissue iron levels in inbred mice. Am J Physiol Gastrointest Liver Physiol. 2017. November 1;313(5):G511–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Bulte JWM. Superparamagnetic iron oxides as MPI tracers: A primer and review of early applications. Adv Drug Deliv Rev. 2019. January 1;138:293–301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Srinivas M, Morel PA, Ernst LA, Laidlaw DH, Ahrens ET. Fluorine-19 MRI for visualization and quantification of cell migration in a diabetes model. Magn Reson Med. 2007;58(4):725–34. [DOI] [PubMed] [Google Scholar]
  • 25.Kadayakkara DK, Ranganathan S, Young W-B, Ahrens ET. Assaying macrophage activity in a murine model of inflammatory bowel disease using fluorine-19 MRI. Lab Investig J Tech Methods Pathol. 2012. April;92(4):636–45. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp info

Supporting Information Table S1a: Correlation coefficients and slopes for R2* vs age, with mice divided by sex and pooled together.

Supporting Information Table S1b. Summary of comparison between slopes of R2* vs age for mice of different strains and diets. A one-way analysis of variance was performed and was significant (p<0.05), then a Tukey-Kramer post-hoc test was used to compare each slope to the others.

Supporting Information Figure S1: Hematocrit for mice from FVB background. Hematocrit was measured for a subset of mice at euthanasia. Mice were grouped by age, strain, and diet. Differences in hematocrit were measured by analysis of variance within age groups, not between ages. The analysis of variance was significant for both analyses (p<0.05). Letters indicate strains that are statistically the same after Tukey’s post-hoc test for multiple comparisons (p<0.05). Dotted line indicates mean hematocrit of 46.6% for 16 week old mice (JAX data sheet-http://jackson.jax.org/rs/444-BUH-304/images/physiological_data_001800.pdf). Numbers at base of bars indicate the number of mice in each group. The hematocrit for the FVBNOG mice on the low-iron diet was significantly lower than the FVBNOG mice on the high iron diet. Mice that were found dead in the cage (presumably due to complications from anemia) or were euthanized before 10 weeks are not included in this measurement so this may be an underestimate of the true effect of the low iron diet on hematocrit.

Supporting Information Figure S2: Additional Histology a) H&E stained section from a 35 wk old FVBGFP-LUC-hiFe mouse with extramedullary hematopoiesis (arrow head). b) An adjacent PPB-stained section where there is no iron deposition corresponding to the EMH. c) H&E stained section from 39 wk old FVBNOG-hiFe mouse with pigment granulomas (arrows). d) An adjacent PPB-stained section where some of the iron staining is co-localized with the pigment granulomas. E) An example of a section that scored a 1 (light spotting throughout) on the PPB scoring system (from a 39 wk old FVBNOG-hiFe mouse). f) An example of a section that scored a 2 (light deposition between lobes) on the PPB scoring system (from a 38 wk old FVBNOG-hiFe mouse). Scale bar is 200 μm.

NIHMS1057029-supplement-Supp_info.pdf (476.9KB, pdf)

RESOURCES