Animal studies were approved by the Institutional Animal Care and Use Committee at Wright State University and conformed to the Guide for the Care and Use of Laboratory Animals (NIH Publication, 8^th edition). Heterozygous breeding pairs of IL-2Rα null (B6;129S4-IL-2ra ^tm1Dw/J) mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) for the establishment of breeding colonies (1).

Antibodies recognizing IL-2Rα were from the following sources: rabbit anti-human polyclonal anti-IL-2Rα antibodies were from Bioss Antibodies, Inc. (Woburn, MA), BosterBio (Pleasanton, CA), and LSBio (Seattle, WA). Mouse anti-human and anti-mouse IL-2Rα, clone 1B5D12, was from Genetex (Irvine, CA). Epitopes for these antibodies are as follows: Bioss aa 258-268; BosterBio aa 39-57; LSBio aa 223-272; Genetex aa 34-139. Rabbit anti-phospho-CD25 (ser268) was from Sigma (St. Louis, MO). Rat anti-mouse CD16/32, rat anti-mouse CD4, and rat anti-mouse IL-2Rα (clone PC61) were from Biolegend (San Diego, CA). Purified mouse IL-2Rα was from SinoBiological (Wayne, PA). Antibodies used for SMC identification were as follows: rabbit anti-human smooth muscle cell alpha actin was from Novus (Littleton, CO). Sheep anti-transgelin (multiple species) was from R&D Systems (Minneapolis, MN). Rabbit anti-human h-caldesmon was from Proteintech (Rosemont, IL). Benzonase^® was from Minnebio (St. Paul, MN). DMEM was from ThermoFisher (Waltham, MA).

Murine VSMC were isolated from thoraco-abdominal aortas by enzymatic digestion based on a protocol from Kwartler, et al. (10). Aortas were dissected free of adipose tissue in situ, removed and rinsed three times in sterile Hanks’ balanced salt solution, then digested at 37°C for 16-18h with a mixture of 0.1 mg/ml collagenase (Sigma), 25 μg/ml trypsin inhibitor (Sigma) and 18 μg/ml elastase (Worthington Biochemical Corp., Lakewood, NJ). Isolated smooth muscle cells and any remaining tissues were then washed and cultured in DMEM with 10% FBS at 37°C and 5% CO₂. Tissue debris was removed the following day.

For serum free conditions, VSMC were cultured in DMEM supplemented with insulin, transferrin, and selenium (ITS; Sigma) (11). Identification of cells as VSMC was confirmed through expression of smoothelin (passage 1-2 only), smooth muscle cell α-actin, transgelin, and h-caldesmon.

Human VSMC were isolated from pieces of aorta using the explant technique (9). Tissues were washed with PBS and all adipose was removed. The artery was then cut into small (approximately 5 mm x 5 mm) pieces and 2-3 pieces per well were placed lumen side down into 6 well tissue culture plates. Pieces were allowed to adhere briefly, then smooth muscle cell media (ScienCell, Carlsbad, CA) supplemented with 10% FBS was carefully added to the wells so as not to dislodge the pieces of aorta. The tissue was maintained at 37°C and SMCs were harvested in 3 - 4 weeks. VSMC were then passaged every 7 - 8 days and used after 1-2 passages.

Offspring of heterozygous breeding pairs of IL-2Rα null (B6;129S4-IL-2ra ^tm1Dw/J) mice were genotyped per Jax mice touchdown protocol (12). Briefly, DNA was extracted from tail clippings and amplified using the recommended primers and touchdown protocol. Primers were as follows: (WT forward CTGTGTCTGTATGACCCACC, WT reverse CAGGAGTTTCCTAAGCAACG; mutant forward CTTGGGTGGAGAGGCTATTC, mutant reverse AGGTGAGATGACAGGAGATC). The WT primer sequences are contained within exon 2.

Quantitative real time PCR was performed, using the above primers, with 15 ng genomic DNA input in triplicate on a 7500 fast real-time PCR system (Applied Biosystems, ThermoFisher Scientific) using PowerUp™ SYBR™ Green Master Mix kit (Applied Biosystems) according to the manufacturer’s instructions. The cycle profile was 2 minutes at 50 °C, 2 minutes at 95 °C, 40 cycles of 3 seconds at 95 °C and 30 seconds at 60 °C. GAPDH was used as an endogenous control. RT-minus and no template controls (NTC) were included for negative control. Data were analyzed using 7500 software v2.3 (Applied Biosystems).

Splenocyte cell suspensions were prepared by mechanical disruption of the spleen followed by filtration to remove large pieces of noncellular debris. The following incubations were all performed on ice. Splenocytes were first incubated with anti-CD16 (TruStain, FcX) for 15 min, then washed and stained with the indicated anti-IL-2Rα antibodies for 30 minutes. Following fixation in 2% paraformaldehyde, intracellular staining was performed by first permeabilizing cells via suspension in ice cold methanol for 30 minutes. Cells were then washed twice in PBS with 2% BSA and stained with anti-IL-2Rα antibodies for 30 minutes. Clone PC 61.5 and anti-IL-2Rα from Genetex were directly conjugated to Cy-5 fluorochromes and anti-IL-2Rα from Boster was used with a Cy5 labeled secondary antibody. A subset of cells was treated with Benzonase^® 250U/10⁶ cells for 50 minutes at 37°C prior to staining. Data were acquired on a Accuri C6 flow cytometer (Bectin Dickinson, Franklin Lakes, NJ) and analyzed using FCS Express 7 (Pasadena, CA).

VSMC were cultured in chamber slides and fixed in ice cold methanol with 2% acetic acid. Fixed VSMC were then washed with PBS and blocked for 1h using Odyssey blocking solution (LI-COR, Lincoln, NE). Blocking solution was removed by washing, and primary antibodies were diluted in TBST (10 mM Tris, pH 8.0, 150 mM NaCl, 0.5% Tween 20) and applied to cells overnight at 4°C. The primary antibody was removed by washing, and the appropriate fluor-conjugated secondary was applied for 1h. Cells were imaged on an EVOS epifluorescent microscope (ThermoFisher) or on a Cytation imaging plate reader (Biotek, Winooski, VT). Images from the Cytation were processed using the flow cytometry software FCS Express 7.

Whole cell lysates from primary VSMCs and Jurkat T cells (ATCC, Manassas, VA) were prepared by lysing and sonicating pelleted cells in radio immuno-precipitation (RIPA) buffer (150 mM NaCl, 25 mM Tris-HCl pH 7.6, 1% Nonidet P40, 1.0% sodium deoxycholate, 0.1% SDS). Splenocytes were separated into membrane, cytoplasmic, and nuclear fractions per manufacturer’s instructions (Cell Fractionation Kit, Cell Signaling Technology, Danvers, MA). Extracts were then separated by SDS-PAGE, using 30 μg protein measured by bicinchoninic acid assay (Pierce, Thermo-Fisher), and transferred to a polyvinylidene difluoride membrane. Some blots, as indicated in figure legends, were assessed for total protein using either stain-free gels (Bio-Rad, Hercules, CA) or the Revert™ 700 Total Protein Stain (LI-COR) prior to blocking. After incubation with 5% nonfat milk in TBST for 60 min, the membrane was incubated with antibodies against IL-2Rα at 4°C for 18 h. Membranes were washed three times for 10 min and incubated with a 1:10,000 dilution of horseradish peroxidase-conjugated anti-rabbit antibodies (Jackson ImmunoResearch, West Grove, PA) at room temperature for 2 hrs. Blots were washed with TBST three times and developed with the ECL system (Luminata Crescendo, Millipore Sigma) according to the manufacturer’s instructions.

The half-life of IL-2Rα was measured using a protocol designed by Morey, et al. (13) in which a traditional pulse-chase experiment was performed using L-azidohomoalanine (AHA), a methionine analog, as the pulsed label. Incorporation of AHA into nascent proteins was detected by a click chemistry based, copper-free strain-promoted alkyne-azide cycloaddition reaction in which a fluorescent cyclooctyne probe (dibenzocyclooctyne-488) bound to incorporated AHA (14, 15).

Specifically, VSMC were washed then cultured in methionine free DMEM with 5% dialyzed FBS containing 50 μM L-azidohomoalanine (AHA) for 48 hours (duration determined by pilot experiments) then cells were returned to DMEM with 10% FBS. VSMC were pelleted at indicated time intervals following this initial pulse and then processed together. Because IL-2Rα is primarily localized to the nucleus in VSMC, we focused on extracting nuclear proteins. Nuclear fractions were isolated following the protocol outlined by Senichkin, et al. (16). Briefly, pelleted cells were incubated in a hypotonic buffer (20 mM Tris-HCl (pH 7.4), 10 mM KCl, 2 mM MgCl₂, 1 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF) for 3 minutes, followed by addition of 0.1% NP-40 (final concentration) for an additional 3 minutes. Following centrifugation, the pellet was incubated in isotonic buffer (20 mM Tris-HCl (pH 7.4), 150 mM KCl, 2 mM MgCl₂, 1 mM EGTA, 0.5 mM DTT, 0.5 mM PMSF) plus 0.1% NP-40 for another 3 minutes, then centrifuged. The resulting nuclear pellet was resuspended in 100 μl Benzonase^®, 100U for 10 minutes at 37°C. RIPA buffer, 400 μl, was then added for 20 minutes at room temperature (RT). Following centrifugation, the supernatant (nuclear soluble fraction) was saved and the insoluble pellet, if present, was washed and resuspended in 100 μl Benzonase^® 100 U. Following a 10-minute digestion at 37°C, the digested lysate was combined with the nuclear soluble fraction for immunoprecipitation, labeling with dibenzocyclooctyne (DBCO)-488, and Western blot analysis.

DBCO labeling and immunoprecipitation were performed as follows. Prepared lysates were heated at 95°C for 10 minutes to denature proteins and increase access to internal AHA-bearing sequences. Iodoacetamide 10 mM, which decreases azide-independent labeling, was added to cooled lysates for 30 minutes at RT (17). Lysates were then immunoprecipitated using anti-IL-2Rα antibody directly conjugated to magnetic beads. Conjugation was performed per manufacturer’s instructions (Click-&-Go, Click Chemistry Tools, Scottsdale, AZ). Following a 1h incubation at RT, beads (now containing bound protein) were washed and incubated with 5 μM DBCO-488 (Click Chemistry Tools) for 30 min at RT or overnight at 4°C. Beads were washed again and labeled protein was eluted with lithium dodecyl sulfate (LDS) containing sample buffer in preparation for SDS-PAGE.

Levels of DBCO-labeled IL-2Rα were normalized using either total protein (Revert™ 700 Total Protein Stain) or total IL-2Rα detected by either the anti-IL-2Rα antibody from BosterBio or anti-phospho-CD25 (ser268). The latter antibody was chosen for its strong signal and suggests that IL-2Rα is active. A linear regression analysis was then performed on normalized levels of labeled IL-2Rα and half-life was calculated via the following equation: t_1/2= ln(2)/slope of decay (13).

Groups were compared using an unpaired, two-tailed t test with Welch’s correction. P values in figures are defined as follows: ns P > 0.05; * P ≤ 0.05; ** P ≤ 0.01; *** P ≤ 0.001; **** P ≤ 0.0001.

To determine how IL-2Rα affects the function of VSMC, we isolated VSMC from IL-2Rα KO and WT mice. WT and KO mice were genotyped using the protocol provided by JAX^® (18). As a first step following isolation and culture, we examined expression of IL-2Rα protein in WT and KO VSMC by indirect immunofluorescence. VSMC isolated from WT mice expressed IL-2Rα primarily in the nucleus, consistent with our recent observations in human VSMC ( Figures 1A, B ). However, IL-2Rα was also detected in the nucleus of VSMC isolated from IL-2Rα KO mice. Similar results were obtained with antibodies from multiple sources. Although most antibodies detected IL-2Rα in the nucleus, two antibodies showed partial (Bioss) or predominantly (Genetex) membrane or cytoplasmic localization ( Figure 1A , Supplementary Figure 1 ). Although we have not established the reason behind this differential detection, it likely relates to the fact that IL-2Rα is heavily glycosylated which may obscure some epitopes. Preadsorption with a known peptide epitope, available for the antibody obtained through BosterBio, abrogated staining ( Figure 1C ).

Membrane IL-2Rα is known to be cleaved by different proteases, at approximately amino acid 192, into a soluble form (19, 20). Because antibodies from LSBio and Bioss recognize epitopes within the C terminus (see Methods for epitopes), and because IL-2Rα from distinct cell fractions are equal in size ( Figure 2C ) this demonstrates that the nuclear form of IL-2Rα is full length and not present through uptake of the soluble form.

IL-2Rα KO mice were generated by replacing exons 2 and 3 of IL-2Rα with a neomycin resistance gene (1). IL-2Rα is known to have multiple splice variants – three have been identified in mouse IL-2Rα and nine in human IL-2Rα (21). To ensure that the IL-2Rα protein detected was not produced by a splice variant, potentially excluding the neomycin resistance insert, we assessed the size of IL-2Rα protein produced by WT and KO VSMC using Western blot analysis. As seen in Figure 1D , WT and IL-2Rα KO VSMC expressed IL-2Rα of the same size, identical to that of human VSMC and Jurkat T cells (9). These results suggest that IL-2Rα KO VSMC express full-length IL-2Rα.

Because IL-2Rα expression and function are mainly studied in T cells, we asked whether lymphocytes from IL-2Rα KO mice express IL-2Rα protein. To this end, we isolated splenocytes from WT and IL-2Rα KO mice and examined IL-2Rα expression by flow cytometry in the presence or absence of stimulation with phytohemagglutinin (PHA) to upregulate membrane IL-2Rα. Since IL-2Rα appears mainly localized to the nucleus in VSMC, we also permeabilized cells to look for nuclear expression. To assess IL-2Rα expression, we used two of the same antibodies used in VSMC, sourced from Genetex and BosterBio (see Figure 1 ).

In non-permeabilized cells, a subset of WT splenocytes expressed IL-2Rα on their cell surface when stimulated with PHA ( Figure 2A ). A small subset of KO cells also expressed IL-2Rα. Upon permeabilization, however, IL-2Rα was detected by in the majority of WT splenocytes (70-80%). IL-2Rα was also detected in permeabilized KO splenocytes, however the frequency was lower than in WT cells, particularly those stained with the anti-IL-2Rα from Genetex. The intensity of staining was also decreased in permeabilized KO vs WT splenocytes ( Figure 2B ).

These results suggested that IL-2Rα could be localized to the nucleus in lymphocytes. To address this question, we separated splenocyte lysates into membrane, cytoplasm, and nuclear fractions. Consistent with our findings in VSMC, IL-2Rα was present in the nucleus ( Figure 2C ).

We then asked whether a commonly used anti-IL-2Rα antibody, rat anti-mouse IL-2Rα clone PC61.5, yielded similar results (22–25). Non-permeabilized WT splenocytes stimulated with PHA yielded strong expression, as expected, and similarly treated KO cells were negative ( Figure 3A ). However, permeabilization yielded only low levels of IL-2Rα in WT cells and KO cells were negative. Given the nuclear localization of IL-2Rα ( Figures 1 , 2 ), we reasoned that a potential association with DNA might be blocking some epitopes. To test this question, we digested permeabilized cells with Benzonase^®, a genetically engineered endonuclease that degrades all forms of DNA and RNA. IL-2Rα was strongly detected in permeabilized WT and KO splenocytes following Benzonase^® digestion, although the level of expression per cell (mean channel fluorescence) was much higher in WT cells ( Figure 3B ). Pre-adsorption of PC61.5 with purified mouse IL-2Rα protein abrogated this staining ( Figure 3A ).

Although prior attempts at using clone PC61.5 to detect IL-2Rα in VSMC yielded poor results, based on the above findings we digested methanol-fixed VSMC with Benzonase^®. Digestion of mouse VSMC with Benzonase^® allowed for the detection of IL-2Rα in the nucleus, consistent with our findings in T cells ( Figure 3C ). In addition, these results suggest that at least a portion of nuclear IL-2Rα binds DNA.

In light of our results demonstrating that both IL-2Rα KO splenocytes and VSMC produce IL-2Rα, we sought to establish its source. As a first step, we asked if the deletion in the IL2RA gene was intact. To this end, we sequenced genomic DNA isolated from WT and KO VSMC. Targeted genome sequencing of the IL2RA gene from KO cells revealed correct deletion of exons 2 and 3 as initially reported by Willerford, et al. (1) ( Figure 4 ).

Given that IL2RA was correctly deleted, we considered other sources of the WT gene. One potential source of the WT IL2RA gene is maternal transfer. Maternal microchimerism, the transfer of cells from mother to fetus, is well established in humans and mice wherein maternal DNA/cells can persist for decades in offspring (26–30). Because IL-2Rα KO mice are infertile, they are generated by mating IL-2Rα heterozygotes; consequently, heterozygous maternal cells are likely present in IL-2Rα KO mice. Maternal microchimerism may therefore represent result a potential source of IL-2Rα in KO offspring of heterozygous females. Our laboratory previously demonstrated that maternal microchimerism was responsible for the presence of IL-2 in IL-2 KO mice, which obscured severe defects in thymic development (29).

Using the IL2RA gene as an indicator of maternal microchimerism, we extracted genomic DNA from IL-2Rα KO tissues and looked for evidence of IL2RA, using primers that amplify a segment within exon 2 that should be absent in KO mice (1). Relative levels of IL2RA were assessed by qPCR of genomic DNA. QPCR of maternal markers in genomic DNA of offspring has been used by other investigators to detect maternal microchimerism (26). As seen in Figure 5A , 6/6 spleens, 6/6 kidneys, and 3/3 hearts from adult KO mice had detectable (RQ > 0 with one of the KO spleens as reference) levels of IL2RA. As further evidence for the presence of maternal cells in offspring of IL-2Rα het breeders, we assessed spleens from IL-2Rα WT offspring of heterozygous parents for the neomycin resistance gene used in the generation of IL-2Rα KO mice. Five of five spleens tested had detectable levels of the NeoR gene ( Figure 5B ). As expected, control spleens from unrelated mice that were not offspring of IL-2Rα het mice had no detectable NeoR gene.

As previously mentioned, IL-2Rα KO mice were generated by replacing exons 2 and 3 of IL-2Rα with a neomycin resistance gene insert as a selectable marker (1). If IL-2Rα KO VSMC contained low numbers of IL-2Rα heterozygous cells, as our results suggest, then treatment with neomycin might eliminate these cells if they are less resistant to neomycin than KO cells. In turn, IL-2Rα protein expression should decrease. To test this hypothesis, we cultured WT and KO VSMC with increasing concentrations of G418 (a neomycin analog). G418 nearly eliminated WT cells at a concentration of 2 mg/ml ( Figure 6A ). However, many VSMC isolated from KO mice survived this treatment. IL-2Rα production by these cells following treatment with G418 was significantly diminished compared to untreated VSMC isolated from IL-2Rα KO mice as shown by both immunofluorescence and Western blot analysis ( Figures 6A-C ). These data suggest that the neomycin resistance protein is expressed and functional in IL-2Rα KO VSMC, and that treatment of KO VSMC with G418 decreased IL-2Rα protein originating from heterozygous cells.

Although our first attempt at clearance of IL-2Rα protein by treatment with G418 significantly decreased IL-2Rα protein expression, we were unable to eliminate IL-2Rα despite ongoing treatment with G418 and/or increased dosage ( Figure 7 ). Upon performing formal kill curves with G418, we observed that the difference in resistance between heterozygous and homozygous cells is probably not large enough to differentially eliminate all heterozygous cells ( Supplementary Figure 2 ).

As part of efforts to identify the source of ongoing IL-2Rα production, we noted that VSMC were heterogeneous in size, and that large cells appeared to express high levels of IL-2Rα protein ( Figure 7A ). Using filters (Pluriselect, Leipzig, Germany), we separated cells into those that passed through a 1 micron filter from the remainder that did or did not pass through a 30 micron filter. We then compared levels of IL2RA DNA amongst these cells and with those that had not received G418. As seen in Figure 7B , larger cells contained relatively more IL2RA DNA than smaller cells, and those without G418 treatment contained the most IL2RA DNA, consistent with our observations of IL-2Rα protein expression ( Figures 6 , 7 ).

Given the above findings, combined with our observation that most VSMC isolated from IL-2Rα KO mice expressed at least low levels of IL-2Rα protein, we asked whether IL-2Rα could be transferred between cells. To this end, we co-cultured IL-2Rα KO VSMC with human VSMC using a transwell insert, in which human VSMC were suspended over KO VSMC and separated by a filter. We chose human VSMC for their robust IL-2Rα protein expression ( Figure 1 ). IL-2Rα was easily detectable in IL-2Rα KO VSMC that had been co-cultured with human VSMC, compared to those that were not ( Figure 7C ). These results suggest that IL-2Rα protein is transferred between cells by a means not requiring cell contact, likely through extracellular vesicles such as exosomes or apoptotic bodies (31, 32).

While the previous results suggest that IL-2Rα protein present in KO mice originated from maternal microchimerism, the low levels of IL2RA DNA detected did not seem consistent with the comparatively high levels of protein. One way in which protein levels may be regulated is through their half-life. In a recent report by Chen, et al, a systematic study of half-lives of proteins in human HepG2 cells showed that most proteins had half-lives ranging from 4-14 hours (33).

Given the easily detectable IL-2Rα protein evident in KO mice, we hypothesized that the half-life of IL-2Rα is long, likely days versus hours. To measure the half-life of IL-2Rα, we performed a pulse chase experiment, in which we labeled newly synthesized proteins with a methionine analog, AHA, then assessed their degradation over time through loss of labeled proteins. Incorporated AHA was detected by fluorescently labeled DBCO, which binds the azido group in AHA through strain promoted alkyne-azide cycloaddition, a copper-free click reaction (14). Based on our results showing that IL-2Rα is localized primarily to the nucleus, we focused on the nuclear fraction. IL-2Rα protein was isolated by immunoprecipitation using the anti-IL-2Rα from BosterBio, and then detected with either the anti-IL-2Rα from BosterBio or an anti-IL-2Rα (phospho-ser 268). Levels of DBCO-labeled IL-2Rα, assessed by densitometry, were normalized based on either total protein or total IL-2Rα protein per band. Normalizing with total IL-2Rα protein was sometimes difficult since the DBCO label variably impacted recognition of IL-2Rα ( Supplementary Figure 3 ). Based on 3 separate experiments, the half-life of IL-2Rα was 8.4 ± 2.5 days ( Figure 8 ). These results demonstrate that the half-life of IL-2Rα is longer than most cellular proteins and may explain the discrepancy between levels of IL2RA DNA and protein observed in IL-2Rα KO mice.

Responses of IL-2Rα KO VSMC to sera, evident in routine culture, began to give us insight into the differences between WT and KO VSMC. Doubling times were significantly shorter in KO vs WT VSMC ( Figure 9 ). KO VSMC were also much smaller than WT and took up less DAPI; these differences were heightened upon stimulation with FBS ( Figure 9 ). The latter observations are consistent with the finding that DNA content is directly proportional to cell size (35). In comparing nuclear area and DNA content of WT and KO VSMC, however, we noted that the decrease in DNA content was out of proportion to the decrease in nuclear area. KO nuclei were approximately half the size of WT, but only had 1/6 the amount of DNA. This finding suggests that KO VSMC are hypodiploid, which implies a severe dysregulation of cell proliferation and/or division. Hypodiploid cells are mainly associated with an aggressive form of acute lymphoblastic leukemia (36, 37).