Skin Wounding–Induced Monocyte Expansion in Mice Is Not Abrogated by IL-1 Receptor 1 Deficiency

This article is featured in In This Issue, p.2515

Monocyte/macrophages (Mo/Mp) are required for repair of a variety of tissues via their roles in the inflammatory, proliferative, and remodeling phases of wound healing (1–4). The importance of Mo/Mp in skin wound repair is evident from studies demonstrating impaired wound healing in mice following selective Mp ablation (3, 5). Mo/Mp accumulate in wounds either by egressing from circulation or migrating and proliferating from their local pool (6–9). However, it is unclear whether skin wounding induces signals that increase Mo production (monopoiesis) in bone marrow (BM) and whether such damage-induced monopoiesis contributes to Mo/Mp accumulation that is required for normal healing of skin wounds.

Published reports have provided evidence that IL-1R1 signaling plays an important role in inflammatory responses to wounding. For example, one study showed reduced infiltration of inflammatory cells in Il1r1^−/− mice leading to decreased scar formation in deep wounds but that skin wound closure was not affected (10). In addition, sustained activation of IL-1R1 signaling in IL-1R antagonist (IL-1Ra)–deficient mice exhibited aberrant wound healing in association with chronic inflammation indicating that timely regulation of IL-1R1 signaling is required for normal wound healing (11). Furthermore, our finding of delayed early skin wound healing in mice deficient for NLR family, pyrin domain-containing 3 (NLRP3), which is required for activation and release of IL-1β, indicates a necessity of IL-1β–induced inflammation in the early stages of normal skin wound healing (12). Finally, we found that blocking IL-1R1 signaling in diabetic db/db mice via local administration of neutralizing Ab to wounds or using Il1r1^−/− BM transplantation induces a switch from proinflammatory to healing associated Mp phenotypes and improves diabetic wound healing (13). However, much remains to be learned about the role of IL-1R1 in regulating the inflammatory response in wound healing.

A recent study demonstrated that chronic IL-1β exposure induces myelopoiesis and myeloid lineage-biased progenitor expansion implicating IL-1R1 signaling in myelopoiesis (14). Another study demonstrated that alum adjuvant induces hematopoietic stem and progenitor cell (HSPC) expansion and subsequent emergency granulopoiesis in wild-type (WT) mice, but that this response was blunted in Il1r1^−/− mice (15). Similarly, myocardial infarction-induced hematopoietic stem cell (HSC) expansion and myelopoiesis was blunted in Il1r1^−/− mice, supporting the importance of IL-1R1 signaling in tissue damage–induced myelopoiesis (16). However, whether damage to other tissues, such as skin wounding, induces HSPC and Mo expansion and whether such a response may be regulated by IL-1R1 signaling has not been reported. The present study aims to fill this gap in the literature.

IL-1 receptor 1 knock out mice (Il1r1^−/−) and C57BL/6 WT controls were obtained from The Jackson Laboratory. Experiments were performed on 8–12-wk-old mice. All experimental procedures involving animals were approved by Animal Care Committee at the University of Illinois at Chicago.

Mice were anesthetized with isoflurane, and their dorsum was shaved and cleaned with alcohol swab. Two 8-mm excisional wounds were made on the back of each mouse with a dermal biopsy punch, and wounds were covered with Tegaderm (3M).

Following 2 d of skin wounding, BrdU (B9285; Sigma-Aldrich) was injected i.v. through retro-orbital route at 2 mg per mouse. After 4 h of BrdU injection, mice were sacrificed and BM cells were fixed, permeabilized, treated with DNase-1, and stained with FITC-conjugated anti-BrdU Ab (364103; BioLegend) following surface Ab staining.

Cells from mouse excisional wounds were dissociated using enzymatic digest as described previously (17). For BM cells, femurs were isolated and marrow was flushed with cold FACS buffer.

For stem cell populations, BM single cells were incubated with biotin-conjugated lineage Abs against CD45R, CD3e, Gr1, TER-119, CD11b, CD4, CD8a, CD19, NK1.1, and CD127 (BioLegend) followed by either streptavidin-allophycocyanin-Cy7 (405208; BioLegend), Sca-1-PerCP-Cy5.5 (108124; BioLegend), cKit–Alexa Fluor 647 (105818; BioLegend), Flk2-BV421 (135315; BioLegend), CD150-PE-Dazzle594 (115935; BioLegend), and CD48-PE (103405; BioLegend) or streptavidin-allophycocyanin-Cy7, Sca-1-PerCP-Cy5.5, cKit–Alexa Fluor 647, FcRγ-V450 (560539; BD Horizon), and CD34-PE (128609; BioLegend). For myeloid cells in BM and blood, cells were stained with Abs against Ly6G-BV421 (127628; BioLegend), CD11b-allophycocyanin-Fire750 (101262; BioLegend), CD115-PE (135505; BioLegend), and Ly6C-PerCPCy5.5 (560525; BD Pharmingen). For wound myeloid cells, enzymatically dissociated single cells were stained for Zombie-BV421 (423113; BioLegend), CD45-FITC (103108; BioLegend), Ly6G-PE-CF594 (562700; BD Horizon), CD11b-allophycocyanin-eFluor780 (47-0112; eBioscience), F4/80-PE (123110; BioLegend), and Ly6C-PerCPCy5.5. For BM cells sorting magnetically enriched progenitors, lineage-BM cells were stained for streptavidin-eFluor450 (48-4317; eBioscience), Sca-1-PE-Cy7 (108114; BioLegend), and cKit–Alexa Fluor 647. For assessing Mo differentiation in culture, differentiated cells were stained for CD11b-allophycocyanin-Fire750 and Ly6C-PerCPCy5.5. For wound cell sorting, wound cells isolated by enzymatic digestion were stained for Zombie-BV421 (423113; BioLegend), CD45-FITC (103108; BioLegend), Ly6G-PE-CF594 (562700; BD Horizon), and CD11b-allophycocyanin-eFluor780 (47-0112; eBioscience). Cell sorting was performed on MoFlo Astrios (Beckman Coulter), and cell analyses were done using LSR II Fortessa (Becton Dickenson).

BM plasma was collected as mentioned previously (18). IL-1β protein was measured in BM plasma using ELISA kit (MLB00C; R&D Systems) following manufacturer’s instructions.

Sorted myeloid progenitor (MyP) cells were grown in StemPro-34 medium (Invitrogen) supplemented with penicillin (50 U ml⁻¹)/streptomycin (50 U ml⁻¹) and l-glutamine (2 mM), SCF (25 ng ml⁻¹), and Flt3L (25 ng ml⁻¹) (PeproTech). IL-1β (211-11B; PeproTech) was added at 5 or 25 ng ml⁻¹ and M-CSF (416-ML; R&D Systems) at 25 ng ml⁻¹.

Results are expressed as mean ± SD. Statistical analyses were performed using Prism 7.0 software (GraphPad). Data between two groups were compared using two-way ANOVA, and different time points in the same group were analyzed using one-way ANOVA (Kruskal–Wallis test) in kinetic assays. Two time points or two groups were compared using Mann–Whitney U test. The p values <0.05 were considered as significant (at 95% confidence intervals).

We examined the kinetics of HSC–long term (HSC-LT) and –short term (HSC-ST), multipotent progenitors (MPP) 2 and 3, common MyP, and granulocyte Mp progenitors (19) in the BM of WT and Il1r1^−/− mice following excisional skin wounding. There was no change in the frequency or number of either HSC-LT (Lin⁻Sca-1⁺cKit⁺Flk2⁻CD48⁻CD150⁺) or HSC-ST (Lin⁻Sca-1⁺cKit⁺Flk2⁻CD48⁻CD150⁻) in WT mice in response to wounding (Fig. 1B, 1C). In contrast, Il1r1^−/− mice showed an increase in the number and frequency of HSC-LT at both 3 and 6 d postinjury (Fig. 1B). Similar to WT controls, Il1r1^−/− mice also did not show any change in HSC-ST subsets following wounding.

MPP3 (Lin⁻Sca-1⁺cKit⁺Flk2⁻CD48⁺CD150⁻) are specific progenitors for myeloid lineage cells (19) and were increased in both WT and Il1r1^−/− mice at 3 d after wounding (Fig. 1D). MPP4 (Lin⁻Sca-1⁺cKit⁺Flk2⁺CD150⁻), which are progenitors for lymphoid lineage cells at homeostatic condition but may switch to producing myeloid cells during inflammatory episodes (19), appeared to show a similar trend as MPP3, but this did not reach statistical significance (Fig. 1E). Further analysis of MyP revealed that the frequency of common MyP (Lin⁻Sca-1⁻cKit⁺FcRγ^loCD34⁺) trended lower in both WT and Il1r1^−/− mice at 6 d after wounding, although their number appeared to be unaltered. In addition, granulocyte Mp progenitor (Lin⁻Sca-1⁻cKit⁺FcRγ^hiCD34⁺) and MyP (Lin⁻Sca-1⁻cKit⁺) subsets were not affected by skin injury in either of the strains (Fig. 1G–I). Similarly, LSK (Lin⁻Sca-1⁺cKit⁺) cells were also found not to be altered by injury in either of the strains (Fig. 1J). Altogether, the data indicate that among all HSPC, only MPP3 is expanded in BM in response to skin wounding. Expansion of HSPC and MPP populations was observed following alum administration in mice (15); our MPP3 population would be found in the HSPC population of this previous study based on the markers used. However, whereas the response to alum was abrogated in Il1r1^−/− mice, loss of IL-1R1 signaling did not influence the MPP3 response following skin wounding.

To determine whether skin wounding–induced expansion of MPP3 cells on day 3 postwounding was due to increased proliferation, we assessed BrdU incorporation 2 d postwounding (Fig. 2). Unexpectedly, the wounding-induced increase in MPP3 cells was not associated with increased frequencies of BrdU⁺ MPP3 compared with unwounded controls in either WT or Il1r1^−/− mice (Fig. 2A). This was also the case when BrdU incorporation was assessed on day 3 postwounding in WT mice (data not shown). In contrast, BrdU incorporation into granulocyte Mp progenitor cells was increased in both WT and Il1r1^−/− mice at 2 d following wounding, whereas their numbers were not increased at any time point after wounding (Figs. 1H, 2B). Similar to MPP3, BrdU incorporation into HSC and common MyP cells was unaltered in response to injury (Fig. 2C, 2D).

IL-1β has been reported to serve as a soluble inflammatory mediator that is transported to BM via the circulation from damaged tissue, leading to increased myelopoiesis following myocardial infarction (16). We hypothesized that similar to myocardial infarction, skin wounding may also elevate IL-1β in BM and induce increased monopoiesis. Hence, we measured IL-1β protein in BM plasma of WT mice following skin wounding, and our results showed no difference in IL-1β level in BM of wounded mice as compared with unwounded controls (Fig. 2E).

The majority of wound Mo/Mp are thought to be derived from circulating cells that are in turn produced in the BM (8, 20). Hence, we assessed Mo kinetics in BM postwounding and tested whether IL-1R1 deficiency influenced Mo kinetics. Results showed significant increase in total Mo (Ly6G⁻CD11b⁺CD115⁺) population in WT BM at day 3 postinjury as compared with noninjured, which reverted back almost to noninjured levels at day 6 postinjury (Fig. 3B). This expansion of BM Mo correlates with the kinetics of BM MPP3 following skin injury (Fig. 1D). Skin wounding also increased the number of Ly6C^hi Mo (Ly6G⁻CD11b⁺CD115⁺Ly6C^hi), and this was associated with increased BrdU incorporation; the frequency of BrdU⁺ Ly6C^hi Mo trended higher in the BM of wounded mice after 2 d of injury compared with unwounded controls (Fig. 3E). In contrast, Ly6C^lo Mo (Ly6G⁻CD11b⁺CD115⁺Ly6C^lo) were found to be decreased in WT BM as the healing progressed and reached their lowest level at 6 d after injury (Fig. 3C, 3D). However, the kinetics of all Mo populations as well as BrdU⁺ Ly6C^hi Mo in Il1r1^−/− BM showed a similar pattern when compared with their WT counterparts (Fig. 3B–E). M-CSF has been reported to induce myelopoiesis in WT HSC by upregulating myeloid-associated transcription factor PU.1, thus indicating an important role of M-CSFR signaling in myelopoiesis (21). Hence, we measured surface M-CSFR on HSPC subsets postwounding to examine whether M-CSFR signaling is involved in skin wounding–associated monopoiesis. Results showed a trend in increased surface M-CSFR protein on LSK and granulocyte Mp progenitors in response to wounding, whereas common MyP did not show wound-induced alteration in surface M-CSFR (Supplemental Fig. 1). Interestingly, surface M-CSFR on granulocyte Mp progenitors and common MyP coincided with BrdU incorporation in these subsets (Fig. 2B, 2D, Supplemental Fig. 1). Furthermore, we observed that granulocyte Mp progenitors from type 2 diabetic mice (db/db) display higher M-CSFR in correlation with increased myeloid expansion in db/db mice (data not shown). Together, these data show that skin wounding–induced monopoiesis in BM is not altered by the loss of IL-1R1 signaling but that M-CSFR signaling may be involved.

To investigate whether monopoiesis is affected by deficiency of IL-1R1 signaling, we isolated MyP from noninjured WT and Il1r1^−/− mice and monitored their differentiation into Ly6C^hi Mo in the presence or absence of IL-1β or M-CSF in liquid culture. As anticipated, MyP cultured from WT mice differentiated and expanded significantly after 4 d in the presence of either IL-1β or M-CSF as compared with untreated controls; both the total cell and Ly6C^hi Mo numbers were increased remarkably (Fig. 4). Also as expected, IL-1β–induced differentiation and expansion was abrogated in MyP cultures derived from Il1r1^−/− mice, confirming that IL-1R1 signaling is indeed nonfunctional in those progenitors (Fig. 4). However, M-CSF–induced expansion of MyP cultures was not inhibited by the deficiency in IL-1R1 signaling. In fact, compared with WT controls the expansion of Il1r1^−/− MyP was significantly higher in response to M-CSF (Fig. 4). Together, our data indicate that although IL-1β–induced monopoiesis is dependent on IL-1R1 signaling, M-CSF–induced monopoiesis is not impaired in the absence of IL-R1 signaling.

Unlike BM, skin wounding did not induce an increase in the frequencies of total, Ly6C^hi, or Ly6C^lo Mo in peripheral blood of either WT or Il1r1^−/− mice (Fig. 5B–D). However, there was a trend and significantly higher percentage of total and Ly6C^lo Mo, respectively, in the circulation of Il1r1^−/− compared with WT mice prior to wounding, in the steady-state (Fig. 5B, 5D); Ly6C^hi Mo did not differ between strains (Fig. 5C). In contrast to peripheral blood, Ly6C^lo Mo did not differ in BM of WT versus Il1r1^−/− mice at any time point (Fig. 3D). Thus, it is possible that the absence of IL-1R1 signaling may increase conversion of Ly6hi into Ly6C^lo Mo in the circulation through an as yet undefined mechanism. However, we did not observe a complementary decrease in Ly6C^hi Mo in the circulation that might be expected with such an increase in conversion.

To examine the effect of IL-1R1 deficiency on the early stages of wound healing, we first measured wound closure in WT and Il1r1^−/− mice in digital photographs of excisional skin wounds. Wound area was reduced in both strains of mice over time but remained somewhat larger in Il1r1^−/− compared with WT mice at 3 d after injury (Fig. 6A). At 6 d postinjury, wound area was found to be similar in both strains (Fig. 6A). Thus, wound closure is modestly slower in Il1r1^−/− mice, which catches up with WT as healing progresses.

As leukocytes play critical roles in wound repair, we examined whether the deficiency in IL-1R1 signaling influenced leukocyte infiltration into the wounds. As anticipated, both the frequency and number of total leukocytes (live CD45⁺) increased remarkably in WT wounds at 3 d after injury and remained high through day 6. Il1r1^−/− mice showed a nearly identical increase in CD45⁺ cells, indicating that wound infiltration of total leukocytes is not affected by IL-1R1 deficiency (Fig. 6C). Further, Mo/Mp (live CD45⁺Ly6G⁻CD11b⁺) were also elevated in WT wounds at 3 and 6 d postinjury as compared with noninjured skin. This increment in Mo/Mp populations followed a similar pattern in Il1r1^−/− mice; however, Il1r1^−/− wounds showed a higher frequency of Mo/Mp than WT controls at both time points (Fig. 6D).

Wound repair is regulated by both the number and phenotype of recruited Mo/Mp, and impaired wound healing has been linked with dysregulation of Mo/Mp subsets (22–24). We used CD11b, Ly6C, and F4/80 to identify different Mo and Mp subsets in the wounds by flow cytometry (1, 4, 17) (Fig. 6B). Our data indicate that wound Mo and Mp subsets (both Ly6C^hi and Ly6C^lo, defined in the figure legend) were increased in both strains in a similar fashion (Fig. 6E–H). But notably, among all Mo/Mp populations, only the frequency of Ly6C^lo Mo was found to be significantly higher in Il1r1^−/− wounds at 3 d postinjury (Fig. 6F). Thus, the higher frequency of total wound Mo/Mp found in Il1r1^−/− mice could be attributed to higher Ly6C^lo Mo (Fig. 6D, 6F). In addition, the higher frequency of Ly6C^lo Mo in day 3 wounds was associated with higher baseline circulating Ly6Lo Mo in Il1r1^−/− mice (Fig. 5D). Altogether, these data may indicate that loss of IL-1R1 signaling leads to higher proportion of Ly6C^lo Mo early in wound healing.

Mo/Mp accumulation in skin wounds is thought to be the result of rapid infiltration of circulating Mo, which, in turn, are supplied by BM (8, 20, 25–27). However, the possibility that skin wounding may produce signals that can communicate with BM to increase monopoiesis has not been addressed. The present study reveals that skin wounding induces Mo expansion in BM during early wound healing. Further, we demonstrate that in parallel with Mo, the myeloid-committed multipotent progenitor MPP3 compartment is also expanded in the BM following skin wounding. Importantly, the loss of IL-1R1 did not affect skin wounding–induced expansion of BM Mo and MPP3 or Mo production by MyP ex vivo.

Our findings of Mo expansion in combination with trends of increased Mo proliferation indicates that skin wounding accelerates monopoiesis in BM during early wound healing. In addition, the higher rate of granulocyte Mp progenitor proliferation in response to wounding suggests their role in this increased monopoiesis. Furthermore, although myeloid lineage–committed MPP3 pool size was increased, their proliferation was not affected by skin wounding. One possible explanation for these seemingly disparate findings could be that the typical sequential differentiation of HSPCs to granulocyte Mp progenitors then to Mo is disturbed during periods of urgent Mo need, such as in the case of wound healing (28). During such periods, the high demand of Mo early after wounding may be supplied directly by granulocyte Mp progenitor pool instead of the usual full monopoietic pathway. As a result, the MPP3 pool may increase because of reduced differentiation, although their proliferation is unaltered. In addition, a recent study from our group has shown that although LSK cell number is increased in WT mouse BM at day 3 after hind limb ischemia, the rate of LSK proliferation is maximum at day 1 and comes back to normal at day 3 postischemia (18). A similar MPP3 response might occur in the response to skin wound healing—MPP3 proliferation might increase at an earlier time point than we assessed, for example day 1 postwounding, leading to increased MPP3 numbers on day 3. Further studies are needed to test these speculative explanations for our findings. Nevertheless, it is evident from the current study that skin wounding causes increased monopoiesis in BM, which likely contributes to wound Mo accumulation during the healing process.

Skin wounding releases a variety of damage-associated molecular patterns with ability to activate inflammatory signaling pathways, such as NLRP3 inflammasome and TLRs, causing increased inflammatory cytokine production (29–32). Activated NLRP3 inflammasome triggers cleavage of pro–IL-1β and pro–IL-18 into their active forms, which function through IL-1R1 signaling (33). Thus, the findings of elevated IL-1β and IL-18 at normal skin wounds suggest that IL-1R1 signaling may play important roles in skin wound healing (13, 34–36). In addition, it is known that IL-1β is transported to BM from infarcted tissue and increases myelopoiesis following myocardial infarction (16). Induction of myelopoiesis by IL-1β exposure and inhibition of alum-induced emergency granulopoiesis in mice deficient for IL-1R1 further consolidates critical roles of IL-1R1 signaling in increased myelopoiesis following injury and inflammatory stimuli (14, 15). We hypothesized that IL-1β, being a soluble factor, may induce monopoiesis by traveling to BM during skin wound healing. But, unlike with myocardial infarction, we did not observe increased level of IL-1β in BM at any time points after skin wounding. Besides, identical expansion of BM Mo in both WT and Il1r1^−/− mice during early skin wound healing suggests that IL-1β does not communicate with BM for inducing monopoiesis after skin wounding. In addition, increased differentiation potential of Il1r1^−/− MyP ex vivo in response to M-CSF indicates that monopoiesis is still functional in Il1r1^−/− BM, and in fact M-CSF–mediated monopoiesis of Il1r1^−/− MyP is even higher than their WT counterparts. Thus, it is likely that another signaling pathway is involved in skin wounding–induced monopoiesis. Our results showing trends of an increase in surface M-CSFR on LSK and granulocyte Mp progenitors in BM following skin wounding, indicating that M-CSFR signaling could be associated with skin wound–induced monopoiesis, and we plan to further investigate this link in future studies.

The current study showed no significant difference in wound closure between WT and Il1r1^−/− mice, which is correlated with similar monopoiesis in the BM of these strains. In contrast, our previous studies have shown sustained IL-1β impairs healing of wounds in diabetic mice and that neutralization of this response by blocking IL-1R1 and NLRP3 inflammasome signaling pathways improves diabetic wound healing (13, 37). Together, these suggest that although IL-1R1 signaling does not affect skin wound healing during normal inflammatory conditions, its dysregulation can promote chronic inflammation and impair wound healing. Whether diabetes-associated chronic inflammation impacts monopoiesis and plays important roles in impaired diabetic wound healing awaits further investigation.

This study focused on the role of IL-1R1 in regulating monopoiesis during skin wound healing based on previous studies that identified such regulation following alum administration or myocardial infarction (15, 16). One limitation of our data is that, although they show definitively that IL-1R1 is not required for MPP3 and Mo expansion in BM induced by skin wounding, the pathways that regulate Mo expansion during skin wound healing will be investigated in future studies. Another limitation of our study is that we have not examined a wider range of kinetics to fully elucidate the dynamics of HSPC proliferation and differentiation, as well as Mo mobilization and death; these processes also await further study.

In short, our data show that skin wounding induces expansion of the myeloid-committed MPP3 compartment in association with increased monopoiesis in the BM during the early stages of wound healing. Further, contrary to alum adjuvant and myocardial infarction-induced myelopoiesis (15, 16), skin wound–induced Mo expansion is not affected by IL-1R1 deficiency.

We thank Dr. Giamila Fantuzzi for input on experimental design and on aspects of the presentation of this manuscript.

The authors have no financial conflicts of interest.