The effect of morphine on rat microglial phagocytic activity: an in vitro study of brain region-, plating density-, sex-, morphine concentration-, and receptor-dependency

Opioids have long been used for clinical pain management, but also have addictive properties that have contributed to the ongoing opioid epidemic. While opioid activation of opioid receptors is well known to contribute to reward and reinforcement, data now also suggest that opioid activation of immune signaling via toll-like receptor 4 (TLR4) may also play a role in addiction-like processes. TLR4 expression is enriched in immune cells, and in the nervous system is primarily expressed in microglia. Microglial phagocytosis is important for developmental, homeostatic, and pathological processes. To examine how morphine impacts microglial phagocytosis, we isolated microglia from adult male and female rat cortex and striatum and plated them in vitro at 10,000 (10K) or 50,000 cells/well densities. Microglia were incubated with neutral fluorescent microbeads to stimulate phagocytosis in the presence of one of four morphine concentrations. We found that the brain region from which microglia are isolated and plating density, but not morphine concentration, impact cell survival in vitro. We found that 10- 12M morphine, but not higher concentrations, increases phagocytosis in striatal microglia in vitro independent of sex and plating density, while 10-12M morphine increased phagocytosis in cortical microglia in vitro independent of sex, but contingent on plating density. Finally, we demonstrate that the effect of 10-12M morphine in striatal microglia plated at 10K density is mediated via TLR4, and not µORs. Overall, our data suggest that in rats, a morphine-TLR4 signaling pathway increases phagocytic activity in microglia independent of sex. This may be useful information for better understanding the possible neural outcomes associated with morphine exposures.


INTRODUCTION
Opioids have long been used for clinical pain management, but also have addictive properties including sex, brain region, and regional microglial density. For example, during neonatal 81 development microglia from female mice are more phagocytically active in the hippocampus 82 (Nelson et al., 2017) and mature faster (Hanamsagar et al., 2017) than microglia from male mice. 83 Morphine increases the number of activated microglia in the female rat periaqueductal gray 84 compared to male rats which correlates to morphine analgesia . Morphine 85 also leads to sex specific microglia polarization in the periaqueductal gray . 86 There also appears to be heterogeneity in the transcriptional profile of cortical and striatal

99
We hypothesized that morphine would modulate phagocytic activity in microglia in vitro, but in 100 a culture density-, concentration-, region-, and sex-dependent manner. Specifically, we predicted 101 microglia would be more phagocytically active, and thus less responsive to the suppressive 102 effects of morphine if they are (1) plated at higher densities relative to lower densities, (2) from 103 striatum relative to cortex, and (3) from females relative to males. To test this, we isolated 104 microglia from adult male and female rat cortex and striatum and plated them in vitro at a 105 density of 10,000 or 50,000 cells (10K or 50K hereafter). Microglia were incubated with neutral 106 fluorescent microbeads to stimulate phagocytosis in the presence of one of four morphine 107 concentrations ( Fig. 1). Unexpectedly, we found that we could not directly compare cortical and 108 striatal microglia, nor the two culture densities, as region and density, but not morphine 109 concentration, impacted cell survival over the course of the experiment. When we separated 110 region and density conditions, we observed that 10 -12 M morphine, our lowest tested 111 concentration, increased phagocytic activity in 10K striatal, 50K striatal, and 50K cortical cells, 112 but not 10K cortical cells. The activating effect of 10 -12 M morphine on 10K striatal, but not 10K 113 cortical cells, was replicated in separate studies, and we further provide evidence that TLR4s, 114 and not µORs is the primary receptor mediating this effect. Overall, our data suggest that a 115 morphine-TLR4 signaling pathway increases phagocytic activity in microglia independent of 116 sex, which is useful information for better understanding the possible neural outcomes associated 117 with morphine exposures.

121
Young adult (postnatal day (P)70-P90) male and female Sprague Dawley rats were purchased 122 (Envigo) and group-housed with ad libitum access to food and water for at least 2 weeks.

123
Conditions were kept at 23°C on a 12:12 light:dark cycle (lights on at 07:00) and cages were 124 changed twice a week. Experiments and animal care were approved by the Institutional Animal 125 Care and Use Committee at Albany Medical College.

127
Tissue collection and microglial isolations 128 Animals were euthanized by CO2 anesthesia and exsanguination. Saline perfusion was performed 129 before animals were decapitated and brains extracted. Each brain had cortex and striatum regions 130 grossly dissected and separated for mechanical and enzymatic breakdown. Brain regions were 131 minced using a razor into a gelatinous consistency while on a petri dish resting on wet ice. and CD11b+ cells positively selected through a magnetic column to enrich for microglial cells. 136 We will refer to CD11b+ cells hereafter as microglia. Microglia comprised on average 28.5% of 137 live cells in pre-isolation samples, and 72.3% of live cells after CD11b+ isolation 138 (Supplemental Fig. 1).

140
Culture plating 141 Microglia were incubated in media (Dubelco's Modified Eagle's Medium with 1µM forskolin 142 and 1% N2 media, 1% L-glutamine, 1% sodium pyruvate, 1% Pen-Strep by volume (v/v)) at 143 37 °C/5% CO2 in chambered slides (Thermo Fisher Scientific Nunc Lab-Tek II CC 2 Chamber 144 Slide, Z734853) at either 10K microglia (10,000 microglia/well or 50,000 microglia/mL) and 145 50K microglia (50,000 microglia/well or 250,000 microglia/mL) density. Microglia were 146 counted using trypan blue (live-dead assay) on an automated cell counter (TC20 Automated Cell 147 counter, Biorad, 508BR08836).     The endpoint that was statistically analyzed was the amount of phagocytosed beads measured as 224 bead fluorescent intensity within microglia, which represents microbead internalization. Outliers 225 were identified with log transformed data due to heteroskedasticity in Graphpad 9.4.1 using the 226 ROUT method (Q=1%); a single outlier was removed from the data presented in Fig. 4, which is  10 -12 M morphine increases phagocytic activity in cortical microglia plated at 50K, but not 254 10K cells per well, and in striatal microglia at either density, independent of sex 255 We could not directly compare cortex and striatum, or 10K and 50K densities due to their 256 differential regulation of microglial survival in vitro (Fig. 2). Thus, we next sought to determine 257 whether sex or morphine concentration would play a role in phagocytic activity in 10K cortex, 258 10K striatum, 50K cortex, and 50K striatum conditions. Microglia in each condition, and from 259 male or female rats, were concurrently incubated with fluorescent microbeads and either 0M, 10 -  33=1.325, p=0.283; Fig. 3). Specifically, 264 10 -12 M morphine increased phagocytic activity relative to all other concentrations in 10K 265 Striatum and 50K striatum groups and increased phagocytic activity relative to 0M and 10 -4 M 266 morphine in 50K cortex groups. There was no significant effect of sex, nor a significant 267 interaction between sex and drug concentration, in any condition.

Morphine-induced increase in phagocytic activity in striatal microglia plated at 10K
270 density is TLR4-, but not µOR-dependent. 271 We next wanted to determine whether 10 -12 M morphine was increasing phagocytosis via TLR4-272 or µOR-dependent mechanisms. Pilot studies using DAMGO as an opioid receptor agonist and 273 naltrexone as a µOR antagonist indicated that 100µM naltrexone would be effective in 274 precluding µOR-dependent changes in microglial phagocytic activity (Supplemental Fig. 2A). 275 Pilot studies using LPS as a TLR4 agonist and RS-LPS as a TLR4 antagonist indicated that 276 1µg/mL RS-LPS would be effective in precluding TLR4-dependent changes in microglial 277 phagocytic activity (Supplemental Fig. 2B). Thus, we incubated cortical and striatal microglia 278 plated at 10K densities with either 100µM naltrexone (or vehicle), 1µg/mL RS-LPS (or vehicle), 279 the combined inhibitors, or combined vehicles for 30 mins. Then, we added a solution containing 280 Fluorescent microbeads, 10 -12 M morphine (final concentration), and the respective inhibitors to 281 maintain their final concentrations, for an additional 90 mins. There were 8 total treatment 282 groups, and because we found no significant sex difference in the effects of morphine on 283 phagocytosis in the studies depicted in Fig. 3, we combined the data from male and female rats. Herein, we demonstrate that the brain region from which microglia are isolated and plating 300 density both impact cell survival in vitro (Fig. 2). We found that 10 -12 M morphine, but not higher 301 concentrations, increases phagocytosis in striatal microglia in vitro independent of sex and 302 plating density (Fig. 3B, D). 10 -12 M morphine also increased phagocytosis in cortical microglia 303 in vitro independent of sex, but this was contingent on a plating density of 50K (versus 10K; Fig.   304 3A, C). This effect was reliable, as an independent replication of the effect of 10 -12 M morphine 305 on 10K cortical and striatal microglia also showed no change in phagocytosis, and increased 306 phagocytosis, respectively (Fig. 4A). Finally, we demonstrate that the effect of 10 -12 M morphine 307 in 10K striatal microglia is mediated via TLR4, and not µORs (Fig. 4B).

309
We initially set out to directly compare the effects of culture density, region, sex, and morphine 310 concentration, as well as their interactions, on microglial phagocytic activity. However, several 311 unexpected findings limited our ability to directly compare the effects of these variables 312 concurrently. We predicted that a higher culture density (50K) would increase phagocytic 313 activity relative to lower (10K) culture density. However, we found that microglia plated at 10K 314 densities did not survive as well as microglia plated at 50K densities, and thus it was not 315 appropriate to make this direct statistical comparison (Fig. 2). This is important to consider when 316 reconciling the literature, which include highly variable culture densities with density rarely 317 included as a variable. In two reports in which microglial phagocytosis is increased by morphine, 318 the plating density was not reported in one (as far as we could find) and was 30,000 cells/well in . Although we observed that morphine generally increased phagocytosis in 50K and 10K conditions, this did not reach statistical 323 significance in cortical microglia plated at 10K density. It is thus possible that cell viability, 324 perhaps in interaction with region examined, may play a role in these disparate conclusions. 325 Finally, although we used an artificially dramatic difference in microglial density to test our 326 hypothesis, it is worth noting that microglia density does vary across the brain, particularly in the  We predicted that striatal microglia would be more phagocytically active than cortical microglia.  (Fig. 2). Although this was a main effect, not an interaction between density and region, 341 visual inspection of the data indicate that this effect is driven primarily by worse survival 342 outcomes in cortical microglia plated at 10K density. We propose that the beneficial effects of

355
We predicted that microglia from female rats would be more phagocytically active than 356 microglia from male rats. To our surprise, we did not observe any significant effects of sex, 357 neither main effects nor interactions, in any of our experiments (Fig. 3). This may be due to the 358 rats being young adults, past the developmental plasticity that might be driving sex differences in We hypothesized that morphine would decrease phagocytosis in a dose-dependent manner.

376
Morphine penetrance into the brain is dependent on concentration, route of administration, increased phagocytosis in 10K Striatum, 50K Cortex, and 50K Striatum conditions (Fig. 3B-D). 380 There was no significant effect of morphine in 10K Cortex conditions (Fig. 3A). It is possible 381 that lower survival in this condition is resulting to the null effect, but visual inspection of the data 382 indicates that story is likely to be more complicated. While there is no statistically significant 383 effect of morphine on cortical microglia plated at 10K density, there is a rather clear pattern of Morphine is documented to activate both TLR4 and µOR signaling to mediate its effects. Thus, 399 in our final set of studies, we tested whether increased phagocytosis in 10 -12 M morphine treated 400 striatal microglia plated at 10K density required TLR4s, µORs, or both. The data indicate that 401 TLR4, but not µORs, is important for this effect (Fig. 4). While the expression of TLR4 is  Overall, our data suggest that a morphine-TLR4 signaling pathway increases phagocytic activity 456 in microglia independent of sex, which is useful information for better understanding the 457 possible neural outcomes associated with morphine exposures.  The percent change in Trypan-Blue indicated live cells from pre-plating to post-experiment was 540 calculated for striatal and cortical microglia plated at 10K or 50K densities and exposed to  581  582  583  584  585  586  587  588  589  590  591  592  593  594  595  596  597  598  599  600  601  602  603  604  605  606  607  608  609  610  611  612  613  614  615  616  617  618  619  620  621  622  623  624   625 Supp. Figure 1: CD11b magnetic isolation enriches for microglia. 626 Microglia were enriched from single-cell suspension of rat brain tissue using Miltenyi Biotec  658  659  660  661  662  663  664  665  666  667  668  669  670  671  672  673