Hematogenous Macrophages Contribute to Fibrotic Scar Formation After Optic Nerve Crush

Abstract

Although glial scar formation has been extensively studied after optic nerve injury, the existence and characteristics of traumatic optic nerve fibrotic scar formation have not been previously characterized. Recent evidence suggests infiltrating macrophages are involved in pathological processes after optic nerve crush (ONC), but their role in fibrotic scar formation is unknown. Using wild-type and transgenic mouse models with optic nerve crush injury, we show that macrophages infiltrate and associate with fibroblasts in the traumatic optic nerve lesion fibrotic scar. We dissected the role of hematogenous and resident macrophages, labeled with Dil liposomes intravenously administered, and observed that hematogenous macrophages (Dil⁺ cells) specifically accumulate in the center of traumatic fibrotic scar while Iba-1⁺ cells reside predominantly at the margins of optic nerve fibrotic scar. Depletion of hematogenous macrophages results in reduced fibroblast density and decreased extracellular matrix deposition within the fibrotic scar area following ONC. However, retinal ganglion cell degeneration and function loss after optic nerve crush remain unaffected after hematogenous macrophage depletion. We present new and previously not characterized evidence that hematogenous macrophages are selectively recruited into the fibrotic core of the optic nerve crush site and critical for this fibrotic scar formation.

Introduction

Traumatic optic neuropathy (TON) is an uncommon but devastating cause of permanent vision loss. TON is typically associated with degeneration of retinal ganglion cells (RGCs) secondary to primary trauma to their axons, a cascade of cellular events associated with inflammatory reactions, and scarring associated with breakdown of the blood–brain barrier [1, 2]. Optic nerve regeneration failure is a critical impediment associated with this repair from this acute damage, which impedes or prevents visual recovery [3]. Unfortunately, no effective treatment for optic nerve trauma exists since we lack a thorough understanding of the basic pathophysiologic and molecular mechanisms for TON and neuro-regeneration.

Studies with spinal cord injury (SCI) reveal that two types of tissue scarring are formed from trauma to the spinal cord: (1) a glial scar that consists of reactive astrocytes, reactive microglia, and glial precursor cells and (2) a fibrotic scar formed by fibroblasts which originate from a specific subset of perivascular cells which gradually invade and reside within the traumatic lesion site [4, 5]. The fibrotic scar is believed to increase the density of the lesion core and creates a physical barrier which influences axon regrowth potential [2].

Therefore, we were interested in whether fibrotic scar formation occurs in optic nerve crush, which is a type of central nervous system (CNS) injury, and is similar to crush injury for the spinal cord. We observed that mature fibrotic scar forms at the crush site approximately 10 days after ONC. This ONC fibrotic scar is composed of multiple cell types, predominantly Col1α1⁺ cells and infiltrated with macrophages and microglia [6]. We observe that CD68⁺ cells and Iba-1⁺ cells accumulate at the fibrotic scar of the ONC site and were not co-labeled with Col1α1⁺ cells, which suggests that peripheral monocyte-derived macrophages and resident microglia contribute to the formation of the optic nerve scar after ONC. Infiltrating macrophages may stimulate fibroblast survival, proliferation, myofibroblast activation, collagen production, and transcriptional activation of pro-fibrotic cytokines, including TGF-β1, PDGF, IL-6, and IL-13 [7]. In liver injury, reduced scarring is associated with decreased myofibroblast infiltration after macrophage depletion [8]. However, the origin of the macrophages residing in the optic nerve crush lesion site and the effects of macrophages on the fibrotic scarring process at the optic nerve injury site are not known.

Clodronate liposome treatment depletes macrophage subsets in various organs. Macrophage-depleting liposomes have been used to explore the effect of macrophages on virus-induced central demyelination [9], spinal cord injury [10], B lymphocyte malignancies and autoimmune disease [11], and atherosclerosis [12]. Some studies also use clodronate liposomes to investigate the role of retinal microglia and vitreous macrophages on retinal capillary neovascularization in ocular ischemic retinopathy mice [13, 14].

In this present study, we describe and characterize the fibrotic scar which forms after traumatic optic neuropathy induced by optic nerve crush. We investigate the composition of monocyte cells at the site of optic nerve injury and explore the effect of macrophage depletion on macrophage accumulation and fibrotic (versus glial) scar formation at the optic nerve injury site, which has important consequences for exploring therapeutic strategies for axonal regeneration and visual recovery after traumatic optic nerve injury.

Method

Mice

Animal procedures were performed in accordance with the University of Miami IACUC and NIH guidelines. Col1α1-GFP mice used have been previously described [15]. Col1α1-GFP mice and wild-type (WT) mice used in our experiments were 8 to 12 weeks old. All the male and female mice were of a C57BL/6 genetic background. Mice were bred and maintained in the University of Miami animal facility and housed under standard conditions of temperature and humidity with a 12-h light/dark cycle and free access to food and water. For all surgical procedures, mice were anesthetized with 100 mg/kg ketamine and 15 mg/kg xylazine intraperitoneally, and the eyes were topically anesthetized using 0.5% proparacaine hydrochloride. Antibiotic ocular ointment containing erythromycin was applied post-operatively to protect the cornea and conjunctival and to prevent infection.

Optic Nerve Crush

Optic nerve crush (ONC) injury was performed as described previously [16]. Eight- to 12-week-old mice were anesthetized and monitored until loss of response to toe-pinch. Under a binocular operating microscope, the superior conjunctiva of the left eye of topical-anesthetized mice was incised and a peritomy performed. The superior ocular muscles were reflected aside, and the optic nerve was exposed at its exit from the eye globe. Dumont #5 forceps (FST) were used to crush the optic nerve approximately 1–2 mm behind the optic disk without damaging retinal vessels or the blood supply.

Macrophage Labeling and Depletion

Liposome administration was performed as previously described [10]. Fluoroliposome-Dil liposomes (50 mg/kg, CLD-8911, Encapula Nanosciences) were injected intravenously via the retro-orbital sinus to label hematogenous macrophages at different time points after ONC. For the 7-day end point, liposomes were injected on days 1 and 4 after ONC; for the 10-day end point, liposomes were injected at 1, 4, and 7 days after ONC; for the 14-day end point, liposomes were injected at 1, 5, and 10 days after ONC. For each group, at least 3 mice were assessed. For macrophage depletion, liposome-encapsulated clodronate (50 mg/kg, CLD-8911, Encapula Nanosciences) was administrated using the same method 3 days prior to ONC and on days 1, 4, and 7 after ONC. Phosphate buffered saline (PBS) liposomes (Encapsome) were used as vehicle controls and injected in the same manner. Mice were sacrificed on day 10 to histologically assess the optic nerve crush lesion and macrophage depletion (Fig. 1).

Fig. 1.

Summary of experimental procedures, treatment groups, and time points for experimental conditions. (a) Experimental timeline for testing the dynamics of hematogenous macrophage infiltration of the fibrotic scar at 7, 10, and 14 days after ONC with Dil liposome sinusvenous injection. (b) Time points of clodronate treatment for macrophage depletion. To identify the efficacy of macrophage depletion, Dil liposome injection is injected on day 5 following ONC. ONC, optic nerve crush

Histology and Immunofluorescence

Mice were anesthetized and perfused transcardially with PBS followed by 4% paraformaldehyde (PFA) in PBS for 5 min to fix tissue. The optic nerves were dissected and placed in 4% PFA in PBS overnight and incubated in 30% sucrose overnight (4 °C). Spleens were fixed for 2 days and incubated in 30% sucrose for 2 days. For histological sectioning, optic nerve samples and spleens were embedded in O.C.T. compound (Tissue-Tek, Sakura Finetek, USA) and serial longitudinal Sects. (10 μm and 7 μm, respectively) were cut using a cryostat and thaw mounted onto Platinum Line Microscope Slides (Mercedes Medical, Germany) and stored at − 20 °C until use.

For immunostaining, non-specific protein binding inhibition was performed with Rodent Block M (Biocare Medical) for 20 min at room temperature. Sections were then incubated with primary antibodies diluted in PBS with 0.5% Triton-X at 4 °C overnight in a humidified box (Table 1). After washing in PBS 3 times for 5 min each, sections were incubated with species-specific fluorescent secondary antibodies for 1 h at room temperature. Control sections were incubated with secondary antibody alone. Finally, sections were cover-slipped with VectaShield (Vector Laboratories) fluorescent mounting medium containing DAPI (Vector Laboratories H-1200). Imaging was performed with a Leica TSL AOBS SP5 confocal microscope (Leica Microsystems, Exton, PA).

Table 1.

Primary antibodies used in the study

Marker (Species)	Dilution	Source
CD3 rabbit	1:100	Abcam, ab5690
CD68 rat	1:400	Abcam, ab53444
Collagen I rabbit	1:200	Abcam, ab34710
F4/80 rat	1:100	Abcam, ab6640
Fibronectin rabbit	1:200	Abcam, ab2413
GFP chicken	1:200	Abcam, ab13970
GFAP goat	1:200	Abcam, ab53554
GFAP rabbit	1:200	Abcam, ab7260
Iba1 rabbit	1:400	Wako, 019–19,741
RBPMS guinea pig	1:200	Phosphosolutions 1832

CD3, cluster of differentiation 3; CD68, cluster of differentiation 68; GFP, green fluorescent protein; GFAP, glial fibrillary acidic protein; Iba-1, ionized calcium binding adaptor molecule 1; RBPMS, RNA-binding protein with multiple splicing

Quantification of Ganglion Cell Layer (GCL) Neurons in the Retina

Whole mount retinas were harvested on day 10 after ONC. Eyeballs were fixed with 4% PFA in 1 × PBS solution for 1 h at room temperature. To create retinal flat mounts, the cornea and crystalline lens were removed, and eyecups were fixed in 4% PFA for another 2 h. The entire retina was then carefully dissected from the eyecup and immersed in 30% sucrose overnight (4 °C). Retinas were permeabilized with 0.5% Triton X-100 in 1 × PBS for 1 h, then blocked to minimize non-specific antibody binding with Rodent Block M for 1 h. Whole mount retinas were incubated overnight at 4 °C in guinea pig polyclonal anti RNA-binding protein with multiple splicing (RBPMS) antibody (phosphosolutions 1832, 1:200) diluted in 1 × PBS with 0.1% Triton-X (pH = 7.4). After 3 washes with PBS, the retinas were treated with the secondary donkey anti-guinea pig antibody for 2 h at room temperature in PBS. Flat-mounted retinas were washed in PBS 3 times for 15 min each, then placed on glass slides (with the RGC layer facing up).

Four radial cuts were made from the edge to the equator of the retina to create a flat retinal wholemount. Retinas were treated with VectaShield fluorescent mounting medium containing DAPI and cover-slipped. In each retinal quadrant, 2 discontinuous images were taken from the central region of the retina to the periphery using a Leica TSL AOBS SP5 confocal microscope (20 × objective, Leica Microsystems, Exton, PA). The number of RBPMS-positive RGCs in each image was quantified with ImageJ software and averaged and normalized to 1 mm².

Pattern Electroretinogram

Pattern electroretinogram (PERG) recording was performed as described by Porciatti [17]. During PERG recording, mice anesthetized with intra-peritoneal ketamine/xylazine were placed on a feedback-controlled heating pad (TCAT-2LV; Physitemp Instruments, Inc. Clifton, NJ) to maintain a constant body temperature at 37 °C. For PERG recording, a semicircular silver loop electrode was placed on the topically anesthetized cornea. Reference and ground electrodes were placed subcutaneously on the back of the head and the base of the tail, respectively. Anesthetized mice received a stimulus of contrast bars (field area, 69.4° × 63.4°; mean luminance, 50 cd/m2; spatial frequency, 0.05 cycles/deg; contrast, 98%; temporal frequency, 1HZ) at a distance of 20 cm. Three independent test trials, composed of 300 measurements per test trial, were recorded for each eye and the data was processed using Sigmaplot (version 11.2; Systat Software, Inc., San Jose, CA). The peak-to-trough amplitude in a time window of 50–300 ms was automatically measured and analyzed by the PERG software. Human bias and recording noise were excluded by the PERG software algorithm [17]. PERG responses were superimposed automatically to assess waveform consistency and then averaged. The major positive (P1) and negative waves (N2), the sum of their absolute values (peak-to-trough amplitude), and the peak latency of the major positive wave (P1) were automatically analyzed, calculated, and graphed using MATLAB software (MathWorks).

Quantification

For each group of immunostaining sections, immunofluorescence signal capture parameters were identical on the confocal microscope for all immunostains. Quantification of immunohistochemical images were performed by two unbiased masked observers using ImageJ software. To quantify fibroblast density at the injury site after clodronate treatment, immunoreactivity for Collagen I was determined by thresholding above background level to the same value and calculating the area covered by the threshold regions using image J and normalized to 1 mm².

To quantify CD68⁺ cell density at the optic nerve injury site, CD68⁺ cells accompanied with DAPI were counted in the area of a GFAP-negative region. The number of CD68⁺ cells in each section was normalized to the area of the GFAP-negative region, and for each animal, the counts from at least three sections were averaged.

For each animal, the optic nerve fibrotic scar area was measured by strictly tracing the GFAP-positive margin near the GFAP-negative regions using ImageJ software and averaged from each section. The cell density in the fibrotic scar was calculated according to the DAPI signal in the GFAP negative regions.

Statistical Analysis

All data are represented as mean ± standard error (n ≥ 4). Statistical significance was assessed using unpaired two-tailed Student’s t test or one-way ANOVA with Tukey’s post hoc test performed using GraphPad Prism software v6.0 (GraphPad Software, Inc., La Jolla, CA, USA). One-way ANOVA with Tukey’s post hoc test was used in the comparison of the density of CD68⁺ macrophages and immunoreactivity of collagen I staining in fibrotic scar area at three time points and the effect of macrophage depletion on RGC survival and P-ERG changes were assessed with three groups.

Results

Hematogenous Macrophages Accumulate Centrally at the Optic Nerve Crush Site

We recently reported that CD68⁺ cells and Iba-1⁺ cells accumulate at the ONC site and to some extent around the ONC site on day 10 after ONC [6]. In our present study, we examined the time course of macrophage infiltration to the optic nerve lesion site following ONC. We used CD68 antibody and ionized calcium binding adaptor molecule 1 (Iba-1) antibody to immunostain for the presence of macrophages/microglia in optic nerves after optic nerve crush. We observed that CD68-positive cells were sparsely present throughout the optic nerve but were highly accumulated within the GFAP-negative regions, which represent the fibrotic scar at the site of optic nerve crush, on days 7, 10, and 14 after ONC (Fig. 2a, g, j). A similar staining pattern was observed using Iba-1 antibody, but as time progressed after ONC, Iba-1⁺ cells gradually accumulated from the center of the fibrotic scar towards the margin of the fibrotic scar (Fig. 2c, f, i, l). CD68 is predominantly expressed by macrophages and activated microglia and Iba-1 is a robust marker for microglial activation in the CNS [18]. Therefore, CD68 is more representative of all the macrophage populations (hematogenous macrophages and microglia) in the ONC fibrotic scar.

Fig. 2.

Macrophages accumulated in the fibrotic scar region after ONC. In non-injured optic nerve, CD68⁺ macrophages (green) and collagen I protein (red) are absent while a small number of Iba-1⁺ cells (grey) are present (a–c). On day 7 following ONC, many CD68⁺ macrophages accumulating in the fibrotic scar (GFAP negative region) (d) are associated with the presence of collagen I protein (e). On day 10, the fibrotic scar is characterized by a dense gathering of CD68⁺ macrophages and fibrosis tissue (collagen I⁺ area) in the crush center (g, h). Similar expression pattern is found on day 14 (j, k). For Iba-1⁺ cells, they populate in the center of fibrotic scar on day 7 and day 10 and gradually reside at the edge of astroglia scar (GFAP-positive region) on day 14 (f, i, l). The density of CD68⁺ macrophages and immunoreactivity of Collagen I staining in fibrotic scar area are calculated during the time course (m, n). Mean values were counted or measured in three randomly selected sections and averaged per optic nerve (at least 4 mice per group). Right side of ON sections is proximal to eyeball. Asterisk indicates significance in one-way ANOVA with Tukey’s post hoc test. * p < 0.05; NS, non-significant by one-way ANOVA with Tukey’s post hoc test. ONC, optic nerve crush. Scale bar, 200 μm

As we previously identified, perivascular Col1α1⁺ cells constitute a major cellular component of the fibrotic scar in Col1α1-GFP reporter mice after ONC [6]. Collagen I antibody was used as a marker for scar-forming cells in our C57/BL6 mice. Collagen I immunostaining demonstrated that on days 7, 10, and 14 following ONC, collagen I is present prominently at the optic nerve crush site that is filled with CD68⁺ macrophages, consistent with our findings in Col1α1-GFP mice models. Similarly, no colocalization was observed between collagen I immunostaining and CD68⁺ cells (Fig. 2e, h, k). CD68 and collagen I expression are absent in the non-crushed optic nerve (Fig. 2a, b).

The macrophage/microglia density in the fibrotic scar (GFAP negative region) was quantified using CD68 as a cellular marker. CD68⁺ cells (macrophages/microglia) increased in a time-dependent manner from 3749 ± 278 to 6068 ± 976 cells/mm² and a significant increase in cell density was quantified between day 7 and day 14 (Fig. 2m, p < 0.05, n = 3–4 for each group). No significant difference in the collagen I fluorescence area was observed among the three groups (Fig. 2n, p > 0.05, n > 5 for each group). These results suggest that CD68⁺ cells are among the major cellular components of fibrotic scarring at the site of the ONC lesion.

To determine the origin of CD68⁺ cells accumulated at the ONC lesion, we use fluorescent Dil liposomes to label monocyte-derived macrophages through intravenous injection. Fluorescent Dil was absent in the non-crushed optic nerve sections although there were a few Iba-1⁺ cells diffusely scattered (Fig. 3a, e, i, m, q). In histological sections of crushed optic nerve (7dpi, 10dpi, and 14dpi), fluorescent Dil liposomes phagocytized by monocytes in the peripheral circulation were observed at the center of the crush site and absent at the peri-lesional regions or distal optic nerve tissues (Fig. 3b–d). Dil⁺ hematogenous macrophages co-localized with the majority of CD68⁺ macrophages at the crush center at three time points (Fig. 3j–l) and were present within the fibrotic scar (GFAP-negative regions) (Fig. 3f–h). Meanwhile, Dil⁺ macrophages were closely localized with collagen I–expressing regions in the optic nerve crush scar center (Fig. 3n–p). On day 7, Dil⁺ macrophages colocalized with Iba-1⁺ microglia at the lesion center, while on day 10 and day 14, most Iba-1⁺ cells surrounded Dil⁺ macrophages (Fig. 3r–t). Therefore, our results demonstrate that hematogenous macrophages are specifically involved in the composition of cellular components of fibrotic scar formation in the ONC, while Iba-1⁺ or CD68⁺ cells around Dil⁺ macrophages may represent reactive resident microglia which infiltrate and migrate not only in the optic nerve crush scar center but also adjacent areas. These results suggest that hematogenous macrophages specifically play a role in the formation of scar tissue in the ONC injury site.

Fig. 3.

Hematogenous macrophages are occupied in the fibrotic scar after ONC. In non-crushed optic nerve, fluorescent Dil (red) and CD68⁺ macrophages (green) are absent while a small number of Iba-1⁺ cells (grey) are present (a, e, i, m, q). On day 7, 10, and 14 after ONC, Dil⁺ hematogenous macrophages are specifically recruited into the fibrotic scar (GFAP – area) (f–h) and co-localized with most CD68⁺ macrophages located in the optic nerve crush injury center (j–l). Dil⁺ macrophages are intimately localized with fibrosis tissue (collagen I⁺ area) in the crush site center (n–p). On day 7, Dil⁺ macrophages are co-localized with Iba-1⁺ cells at the lesion center (r) while most Iba-1⁺ cells surround Dil⁺ cells and are present at the edge of fibrotic scar on day 10 and 14 (s, t). Dil⁺ liposomes were injected on days 1 and 4 for 7-day post-injury group, injected at 1, 4, and 7 days for 10-day post-injury group, and injected at 1, 5, and 10 days for 14-day post-injury group. Right side of ON sections is proximal to eyeball. ONC, optic nerve crush. Scale bar, 200 μm

Hematogenous Macrophage Depletion in the Optic Nerve

In order to test the role of hematogenous macrophages in optic nerve crush fibrotic scar formation, we depleted hematogenous macrophages by injecting mice with systemic clodronate liposome 3 days before ONC. We confirmed macrophage depletion efficiency in peripheral tissue such as the spleen. Ten days after ONC and clodronate treatment, light microscopic examination of the H&E paraffin-stained spleen sections showed that splenic red pulp (margins depicted by arrows) exhibited a marked decrease in extramedullary hematopoiesis compared with that of control PBS-treated mice (Supplemental Fig. 1). Immunohistochemistry revealed that F4/80-expressing phagocytic cells were highly present in the red pulp in the control PBS-treated group (Supplemental Fig. 1c), whereas these cells were nearly absent in the clodronate-treated group (Supplemental Fig. 1d). However, cells staining for CD3 (a T cell marker) are mostly present in the white pulp and unaffected by clodronate treatment (Supplemental Fig. 1e, f). These results demonstrate that clodronate liposome treatment readily and selectively depletes general phagocyte populations in the peripheral tissue but not non-phagocytic cells, such as lymphocytes.

Secondly, we utilized intravenously injected fluorescent Dil liposomes to further ensure the depletion of hematogenous macrophages in the ONC scar. We observed with the same clodronate administration procedures that Dil fluorescence signals were decreased in the fibrotic scar region (GFAP-negative area) compared with those of PBS-treated ONC mice (Supplemental Fig. 1i–l). Similarly, the number of CD68⁺ cells (hematogenous macrophages/microglia) in the ONC lesion and CD68⁺ cells co-localized with Dil signals was significantly reduced in the clodronate-treated group (Supplemental Fig. 1i, j). Clodronate liposome intravenous injection efficiently depletes hematogenous macrophages in the ONC lesion.

Hematogenous Macrophage Depletion Reduces Fibrotic Scar Formation

The optic nerve fibrotic scar matures by approximately day 10 after ONC (Figs. 2 and 3), so we used this time point to investigate the correlation between the depletion of hematogenous macrophages and changes in fibrotic scar formation. Along with a dense accumulation of CD68⁺ cells in the optic nerve crush lesion center (Fig. 4b), collagen I immunofluorescence staining showed well-defined spindle-like or square-like scaffolds in the center of the ONC sites (Fig. 4e) and similar staining patterns were observed with fibronectin (Fig. 4n). This suggests fibroblasts (which are the main cell type producing extracellular matrix) are attracted to the ONC lesion, accumulate in the center of the optic nerve crush site, and participate in fibrotic scar formation.

Fig. 4.

Hematogenous macrophage depletion reduces fibrotic scar formation after ONC in wild-type mice. Representative images of CD68-stained optic nerve tissue (a–c), collagen I–stained optic nerve tissue (d–f), GFAP-stained optic nerve tissue (g–i), and fibronectin-stained optic nerve tissue (m–o) from PBS-treated or clodronate-treated wild-type mice at day 10 after ONC and non-crushed mice. Merged images (j–l) are from CD68, collagen I, and GFAP staining as shown above. (p) Quantification of CD68⁺ cell density in the fibrotic scar area (according to GFAP-negative area) per square millimeter as indicated in (b) and (c). Density of CD68⁺ cells in the fibrotic scar is significantly reduced after macrophage depletion. (q) Quantification of collagen I fluorescence area in the optic nerve sections as represented in (e) and (f). Areas covered by the threshold regions using image J in the 1.6 mm-length area including crush site are measured and normalized by the optic nerve area in this sector. Macrophage depletion reduced collagen I⁺ density. (r) Quantification of cell density (according to DAPI staining) in the fibrotic scar area (according to GFAP-negative area) per square millimeter. Cell densities in the fibrotic scar area are significantly reduced after macrophage depletion. (s) Quantification of GFAP fluorescence negative region on optic nerve sections as represented in (h) and (i). Macrophage depletion enlarged GFAP-negative area. Mean values were counted or measured in three randomly selected sections and averaged per optic nerve (over five mice per group). Right side of ON sections is proximal to eyeball. Asterisks indicate significance in unpaired two-tailed student’s t test. * p < 0.05; ** p < 0.01. Scale bar, 200 μm

With hematogenous macrophage depletion by clodronate administration, the density of CD68⁺ cells in fibrotic scar area was greatly reduced compared with that of the PBS-treated ONC group (from 4935 ± 1091 to 3367 ± 221 cells/mm², p = 0.007, t = 3.470, Fig. 4b, c, p, n > 5 for both groups) and this correlated with a significant reduction of cell density in fibrotic scar area (from 7025 ± 522 to 4389 ± 1341 cells/mm², p = 0.0001, t = 5.207, Fig. 4r, n > 5 for both groups). After macrophage depletion, the fibrous optic nerve scar scaffold was greatly disrupted (Fig. 4f, o). Regions of collagen I fluorescence were measured in a 1.6 mm-length area, which encompasses the optic nerve crush site, and normalized by the optic nerve area in this circumscribed area. The percent of collagen I⁺ expression decreased from 2.9 ± 1.44% to 1.7 ± 0.67% (p = 0.028, t = 2.414, Fig. 4q, n > 5 for both groups) from control to crushed regions of the optic nerve. The average fibrotic area (GFAP-negative area) on each optic nerve section was significantly enlarged after clodronate treatment (0.16 ± 0.06 mm²), compared with that of the PBS-treated group (0.10 ± 0.02 mm², p = 0.0068, t = 3.138, Fig. 4h, i, s, n> 5 for both groups).

We confirmed these results using col1α1-GFP mice, which have GFP expression regulated by the collagen 1α promoter, as a model system to investigate the mechanisms affecting collagen expression during fibrosis [15]. Collagen I–associated fibrosis is illustrated by GFP expression since col1α1-GFP mice have GFP expression regulated by the collagen 1a promoter. A mature and roundly circumscribed encapsulated fibrotic scar (Col1α1-GFP expressing) was typically observed by day 10 following ONC (Fig. 5b), along with marked and simultaneous infiltration of CD68⁺ cells in the GFAP-negative region (Fig. 5e, h). These results are consistent with our data using C57/BL6 mice (Fig. 2). Similarly, the col1α1⁺ scaffold structure of the optic nerve crush fibrotic scar collapsed with hematogenous macrophage depletion in the ONC lesion of col1α1-GFP mice (Fig. 5c, f, i). These results suggest that depletion of macrophages in the ONC site is associated with decreased fibrotic scar formation and reduced optic nerve fibrotic scar density.

Fig. 5.

Verification of effect of hematogenous macrophage depletion in Col1α1-GFP mice after ONC. (a–c) Representative images of GFP-stained optic nerve tissue from PBS-treated or clodronate-treated col1α1-GFP mice at day 10 after ONC and non-crushed mice. (d–f) Representative images of col1α1-GFP and CD68 double-stained optic nerve tissue from PBS-treated or clodronate-treated col1α1-GFP mice at day 10 after ONC and non-crushed mice. (g–i) Representative images of col1α1-GFP and GFAP double-stained optic nerve tissue from PBS-treated or clodronate-treated col1α1-GFP mice at day 10 after ONC and non-crushed mice. Right side of ON sections is proximal to eyeball. Scale bar, 200 μm

Hematogenous Macrophage Depletion Does Not Prevent the Degeneration and Functional Loss of RGCs

Axonal injury in the optic nerve triggers cell death and degeneration of mature RGCs. RGC soma are cut off from their nutrient supply of neurotrophic factors after axonal injury, since survival factors for the RGC are retrogradely transported from the synaptic terminals in the brain through the axons [19]. Therefore, we determined if RGC death or axonal degeneration was decreased by the reduction of fibrotic density in the optic nerve crush fibrotic scar center after clodronate treatment. At day 10 following ONC, RGC numbers (RBPMS-positive cells) in the whole-mounted retinas were dramatically decreased compared with those of the non-crushed eyes, while no significant quantitative difference was measured between PBS-treated and clodronate-treated ONC groups (Fig. 6a, b, n = 6 for ONC 10dpi + PBS group, n = 11 for ONC 10dpi + clodronate group, n = 3 for the control group). In addition, the pattern ERG amplitude of mice at 10 days after ONC sharply dropped and lost normal PERG wave shape compared with that of the uninjured group. However, the PERG waveform of ONC mice treated with clodronate did not improve remarkably (Fig. 6c, d, n ≥ 9 for each group) even though there was decreased fibrotic scar density at the ONC site. Taken together, these data suggest macrophage depletion does not inhibit RGC degeneration or mitigate the function loss of the inner retina after ONC injury.

Fig. 6.

Hematogenous macrophage depletion did not improve RGC survival or preserve RGC function after ONC. (a) Representative images of RBPMS-stained retinal wholemounts from PBS-treated or clodronate-treated wild-type mice at day 10 after ONC. (b) quantification result of RBPMS-positive RGC density normalized to 1 mm². No significant difference was observed between PBS-treated and clodronate-treated ONC groups. (c) Representative images of PERG results from PBS-treated or clodronate-treated wild-type mice on day 10 after ONC and the control mice. (d) Quantification of PERG amplitude from three groups indicated in (c). No significant difference was observed between PBS-treated and clodronate-treated ONC groups. ONC, optic nerve crush; RBPMS, RNA-binding protein with multiple splicing. Asterisk indicates significance in one-way ANOVA with Tukey’s post hoc test. ** p < 0.01; NS, non-significance by one-way ANOVA with Tukey’s post hoc test

Discussion

We demonstrate in this study that activated microglia/macrophages and collagen I⁺ fibroblasts are the major cellular components within the core of the fibrotic scar formed after optic nerve crush–induced traumatic optic neuropathy. Dil⁺ hematogenous macrophages are selectively recruited into the core of the ONC scar site accompanied by the presence of fibrotic scaffold proteins. Depletion of hematogenous macrophages by systemic clodronate liposome administration reduced the density of fibrosis accompanied by enlargement of a GFAP-negative area within the ONC scar center. These results in wild-type C57Bl/6 mice were confirmed with transgenic mice expressing GFP under the collagen1α1 promoter (Col1α1-GFP mice). Interestingly, hematogenous macrophage depletion did not prevent RGC degeneration and functional loss from optic nerve crush injury.

Two distinct origins of macrophages after injury are believed to exist within the monocyte/phagocyte system: (1) microglia that are established prenatally from the yolk sac and (2) tissue-infiltrating monocytes, which are associated with pathological, homeostatic, and inflammatory processes [20]. Impaired recruitment of monocytes or monocyte-derived cells can attenuate the clinical symptoms of autoimmune multiple sclerosis model (EAE) animals [21], but discerning hematogenous monocyte-derived macrophages from resident microglia is difficult.

We used fluoroliposome-Dil liposomes to visualize macrophages from the peripheral hematogenous circulation. Surprisingly, Dil⁺-containing cells were only observed within the ONC lesion core (GFAP-negative area). These cells largely co-localized with CD68⁺ and Iba-1⁺ cells at day 7 (Fig. 3j–l, r–t) and closely correlated with collagen I⁺ staining, which is associated with fibrosis (Fig. 3n–p). These results are consistent with those observed in spinal cord injury [10], suggesting that hematogenous macrophages may play an important role in fibrotic scar formation following ONC injury to the optic nerve. On day 10 and day 14 after ONC with the presence of a mature fibrotic scar, more Iba-1⁺ cells without Dil⁺ staining were observed around the fibrotic scar and distal regions of the ONC site (Fig. 3s, t). Previous studies showed that in most cases, Iba-1 can be used to differentiate microglia from other cell-related populations [18, 22, 23]. To further distinguish microglia from peripheral-derived macrophages, transmembrane protein 119 (Tmem119) antibody has recently been recognized as a microglia-specific marker [24]. Tmem119⁺ cells were mainly located at the periphery and outside of the fibrotic scar (GFAP-negative area) 10 days after ONC, which is similar to the appearance of Iba-1 immunostaining results (Supplemental Fig. 2). Therefore, we speculate that different macrophage subtypes may have distinct roles in ONC scar formation according to their different spatial distribution in scar formation.

Previous studies demonstrate that the development of fibrosis is associated with activation of monocytes and macrophages [25]. In the process of fibrosis formation, infiltrating inflammatory cells release cytokines and stimulate fibroblasts to migrate, proliferate, and secrete de novo extracellular matrix components [26]. However, macrophages from tissue-resident sources and circulating macrophages appear to play different roles in tissue repair [7]. Acute lung injury studies suggest recruited macrophages, which originate post-natal from circulating monocytes and distinct from resident macrophages, have pro-inflammatory functions by producing inflammatory cytokines and increasing glycolytic metabolism [27]. We therefore hypothesize that inhibition of hematogenous macrophage recruitment could reduce fibroblast accumulation in the ONC lesion core.

Our data demonstrate that after clodronate treatment, the density of mature fibrotic scar (distinguished by type I collagen, fibronectin, and col1α1-GFP expression) formed 10 days after ONC was significantly decreased in both wild-type C57/BL6 and transgenic col1α1-GFP mice. Fibrotic scar cell density in the GFAP-negative regions of ONC was significantly decreased as well, suggesting that both cellular components and extracellular matrix deposition were diminished in the ONC center after hematogenous macrophage depletion. However, the ONC lesion area devoid of GFAP-expressing cells was enlarged after clodronate treatment, which is similar to observations in contusion injury to the spinal cord [10]. Previous studies reveal that M1 macrophages (pro-inflammatory) play an important role in the development of astrogliosis following spinal cord injury [28] and a M1 macrophage response is rapidly induced and then maintained at injured spinal cord sites [29]. We cannot rule out the possibility of a delay or decrease in astrocyte response after macrophage depletion in the acute phase of ONC. Therefore, future studies are needed to determine the effect of hematogenous macrophage depletion on gliosis in the acute and chronic phases after ONC. Based on our results, hematogenous macrophage depletion creates a less crowded and dense fibrotic scar lesion environment which theoretically could improve axonal regeneration and functional recovery of the optic nerve lesion in an appropriate cellular environment.

Unfortunately, we did not observe improvements in RGC survival or electrophysiological function after clodronate treatment compared with PBS treatment following ONC injury. Similarly, Sanchez et al. [30] observed that HDAC3 inhibition promoted a shift from M1-like to M2-like (anti-inflammatory/regenerative type) macrophages and reduced the formation of foamy macrophages, but did not lead to improved functional recovery after spinal cord injury. RGC degeneration and axon regeneration rescue after ONC were also not observed with microglia depletion [22]. These studies suggest that interference with macrophage infiltration and activity and structural alterations in the ONC fibrotic scar is not enough to achieve a favorable functional outcome but may be part of combinatorial treatments to increase the possibility of restoration of neuronal connectivity and function around and within the fibrotic scar formed following ONC.

Conclusions

To the best of our knowledge, our study is the first to demonstrate that macrophages, especially hematogenous macrophages, are one of the major cellular components in the traumatic fibrotic lesion core formed after ONC and hematogenous macrophages are associated with fibrotic extracellular matrix deposition. Depletion of hematogenous macrophages reduced the density of fibrosis in the center of the ONC lesion. Further explorations of the molecular mechanisms underlying the role of hematogenous macrophages and combination treatment after optic nerve injury are needed to pursue visual recovery after traumatic optic neuropathy.

Abbreviations

CD3

Cluster of differentiation 3

CD68

Cluster of differentiation 68

CNS

Central nervous system

GCL

Ganglion cell layer

GFP

Green fluorescent protein

GFAP

Glial fibrillary acidic protein

Iba-1

Ionized calcium binding adaptor molecule 1

IL-6

Interleukin-6

IL-13

Interleukin-13

ONC

Optic nerve crush

PBS

Phosphate buffered saline

PDGF

Platelet derived growth factor

P-ERG

Pattern electroretinogram

PFA

Paraformaldehyde

RBPMS

RNA-binding protein with multiple splicing

RGC

Retinal ganglion cells

SCI

Spinal cord injury

TGF-β1

Transforming growth factor-beta1

Tmem119

Transmembrane protein 119

TON

Traumatic optic neuropath

Data Availability

Please contact the corresponding author for data requests.