Data on leukocyte PDZK1 deficiency affecting macrophage apoptosis but not monocyte recruitment, cell proliferation, macrophage abundance or ER stress in atherosclerotic plaques of LDLR deficient mice

PDZK1 (Post-synaptic density protein/Drosophila disc-large protein/Zonula occludens protein containing 1) is an adaptor protein that binds to the HDL receptor, Scavenger receptor class B type I. Leukocyte PDZK1 deficiency in high fat-diet fed LDL receptor knockout mice has been found to increase atherosclerotic necrotic core formation and apoptosis of cells within atherosclerotic plaques. To explore mechanisms that may be involved, we examined the effects of leukocyte PDZK1 deficiency in mice on a number of processes that may impact macrophage abundance within atherosclerotic plaques. We found that leukocyte PDZK1 deficiency in high fat diet fed LDL receptor knockout mice did not affect the abundance of circulating red blood cells, myeloid cells or B- or T-lymphocytes. Leukocyte selective PDZK1 deficiency did not affect the levels of the ER chaperone proteins, detected with an antibody against the KDEL peptide, in macrophages or macrophage abundance, cellular proliferation or monocyte recruitment in atherosclerotic plaques. Leukocyte PDZK1 deficiency in otherwise wild type mice did result in increased sensitivity of macrophages to tunicamycin-induced apoptosis in a peritonitis model. HDL protected wild type macrophages from apoptosis induced by a variety of agents, including the ER stressor tunicamycin, oxidized LDL and exposure to UV irradiation. However, this protection afforded by HDL was lost when macrophages were deficient in PDZK1. HDL did not affect the level of ER stress induction by tunicamycin. Finally, PDZK1 deficiency in macrophages did not affect lipopolysaccharide-mediated induction of markers of M1 polarization. These data, utilizing mouse and cellular models, help to demonstrate that leukocyte PDZK1 plays a role in atherosclerosis by affecting macrophage apoptosis within atherosclerotic plaques.

sensitivity of macrophages to tunicamycin-induced apoptosis in a peritonitis model. HDL protected wild type macrophages from apoptosis induced by a variety of agents, including the ER stressor tunicamycin, oxidized LDL and exposure to UV irradiation. However, this protection afforded by HDL was lost when macrophages were deficient in PDZK1. HDL did not affect the level of ER stress induction by tunicamycin. Finally, PDZK1 deficiency in macrophages did not affect lipopolysaccharide-mediated induction of markers of M1 polarization. These data, utilizing mouse and cellular models, help to demonstrate that leukocyte PDZK1 plays a role in atherosclerosis by affecting macrophage apoptosis within atherosclerotic plaques.  Histological sections of atherosclerotic plaques and primary macrophages from experimental mice were used.

Data source location
Hamilton, Ontario, Canada Data accessibility Data included in this article and is related to articles published

Value of the data
The data presented herein is key to understanding the consequences of inactivating PDZK1 gene expression in bone marrow derived cells on atherosclerosis development.
This data gives insight into mechanisms by which PDZK1 influences atherosclerosis development. This data provides a more thorough understanding of how PDZK1 protects macrophages against apoptosis induced by different stressors.

Effects of bone marrow selective inactivation of PDZK1 on atherosclerotic plaques in ldlr KO mice
To determine the effects of bone marrow selective inactivation of PDZK1 on high fat diet induced atherosclerosis, low density lipoprotein receptor (ldlr) knockout (KO) mice were transplanted with bone marrow (BM) from either pdzk1 KO or corresponding wild type (wt) mice, allowed to recover for 4 weeks and then fed a high fat diet for a further 10 weeks. BM-specific pdzk1 deletion did not significantly affect hematocrits, red blood cell sizes, or proportions of leukocytes that were positive for CD3, B220 or CD11b (Table 1). We detected no differences in the extent of immunostaining with an antibody against the -KDEL endoplasmic reticulum (ER) retention peptide, which detects the major ER chaperones, as a measure of ER stress (Fig. 1A, B, H). Similarly, we detected no differences in macrophage abundance (Mac3 immunostaining (Fig. 1C, D, G)) or in cell proliferation (Ki67 staining; Fig. 2) or monocyte recruitment into plaques (Fig. 3). On the contrary, in a parallel study [1], we detected increased atherosclerotic plaque sizes and increased cell apoptosis within atherosclerotic plaques of ldlr KO mice transplanted with BM from pdzk1 KO donors, and subsequently fed the high fat diet.

BM selective inactivation of PDZK1 increases sensitivity of peritoneal macrophages to ER stress induced apoptosis
To test the effects of BM specific inactivation of PDZK1 on the sensitivity of macrophages to apoptosis, wild type mice were transplanted with BM from either pdzk1 KO or control wt donors, allowed to recover for 8 weeks and then injected i.p. with thioglycollate to induce macrophage recruitment. Three days after thioglycollate injection, mice were injected i.p. with tunicamycin to induce ER stress and apoptosis in peritoneal cells. The next day, peritoneal cells were recovered, immunostained for apoptosis induction using an antibody for cleaved (activated) caspase 3 (CC3) and analyzed by flow cytometry (Fig. 4). We saw no induction of apoptosis by tunicamycin in peritoneal macrophages from mice transplanted with wt BM, but significant induction of apoptosis by tunicamycin in peritoneal macrophages from mice transplanted with pdzk1 KO BM. Furthermore, we detected increased basal apoptosis in mice transplanted with pdzk1 KO BM compared to mice transplanted with wt BM. In the accompanying article [1], we saw similar results for mice with whole body pdzk1 KO compared to wt mice, although in that case, we detected no differences in basal apoptosis.

PDZK1 is required for HDL mediated protection against apoptosis induced by different agents
Peritoneal macrophages were prepared from wt and pdzk1 KO mice and analyzed in culture. Cells were treated with different apoptosis inducing agents, including tunicamycin, oxidized LDL (oxLDL) and exposure to UV irradiation. Apoptosis was measured by CC3 (Fig. 5), Annexin V ( Fig. 6A-I) or TUNEL (Fig. 6J,K) staining. Fig. 5 shows representative images of wt and pdzk1 KO macrophages that were either untreated or treated with tunicamycin in the absence or presence of HDL, prior to detection of apoptosis induction by staining for CC3. Data corresponding to these representative images was quantified and is presented as Fig. 3I in Ref [1]. We detected increased activation of caspase 3 in both wt and pdzk1 KO macrophages treated with tunicamycin alone; furthermore, the extent of caspase 3 activation was suppressed in the presence of HDL in wt but not in pdzk1 KO macrophages. Similarly, oxLDL increased apoptosis, measured as Annexin V cell staining, of both wt and pdzk1 KO macrophages and HDL was able to suppress this in wt but not in pdzk1 KO macrophages ( Fig. 6 A-I). Similar results were obtained when apoptosis was induced by treatment with oxLDL ( Fig. 6J) or exposure of cells to UV irradiation (Fig. 6K) and apoptosis was measured by TUNEL staining. 29.9 7 2.5 (n ¼ 9) 26.5 7 1.6 (n ¼ 11) 0.27 a Statistical analysis was done using the Mann-Whitney rank sum test. b Hematocrit, MCV and RDW was analyzed by Hemavet analysis of whole blood c % CD3 þ , B220 þ and CD11b þ cells were determined by flow cytometry and are expressed as the proportions of total leukocytes.

HDL treatment of macrophages does not prevent tunicamycin mediated induction of markers of ER stress/unfolded protein response
Treatment of macrophages with tunicamycin triggered caspase 3 activation (Fig. 5) and apoptosis [1] and this could be inhibited by HDL treatment of wt macrophages. Because tunicamycin is known to trigger apoptosis as a result of the induction of ER stress and the unfolded protein response [2], we tested if HDL treatment affected the ability of tunicamycin to induce markers of ER stress/unfolded protein response in wt macrophages (Fig. 7). We saw robust induction of the glucose regulated proteins of 94 and 78 kDa (GRP94 and GRP78) protein levels and of the mRNA for the C/EBP homologous protein (CHOP) in cells treated with tunicamycin. HDL treatment, however did not affect the ability of tunicamycin to induce these markers of ER stress/unfolded protein response (Fig. 7).

HDL does not induce STAT3 phosphorylation at the concentrations effective at protecting against apoptosis
In the accompanying paper [1] we report that HDL treatment of peritoneal macrophages from wt but not pdzk1 KO mice induced increased AKT phosphorylation. Because the signal transducer and activator of transcription 3 (STAT3) has been implicated by others [3] in HDL-mediated protection of RAW264.7 macrophages against apoptosis, we tested the effects of treatment of mouse peritoneal macrophages with HDL at 50 mg protein /ml, a concentration effective at protecting against apoptosis (Figs. 5 and 6), on the levels of STAT3 phosphorylation. We saw that treatment with 50 mg/ml HDL did not significantly affect the levels of STAT3 phosphorylation at Y705 (reported to induce STAT3 dimerization [4]) in wt macrophages or in macrophages from pdzk1 KO mice (Fig. 8). We also tested macrophages from akt1 KO or akt2 KO mice. Again, HDL treatment did not increase STAT3 Y705 phosphorylation. However baseline STAT3 Y705 phosphorylation appeared to be increased in akt1 KO and akt2 KO macrophages compared to wt macrophages (Fig. 8).

Pdzk1 KO macrophages do not exhibit evidence of increased necroptosis induction
It has been reported that oxLDL treatment of macrophages can, under certain circumstances, trigger the induction of necroptosis, or programmed necrosis [5]. This involves the phosphorylation of the receptor interacting protein (RIP) kinases RIP1K and RIP3K and of mixed lineage kinase domain like (MLKL), which, upon phosphorylation, inserts into the plasma membrane and oligomerizes to form pores, causing cellular necrosis [6]. We therefore examined the level of phosphorylated MLKL in wt and pdzk1 KO macrophages treated with oxLDL (100 mg protein/ml) for either 8 h or 24 h, corresponding to the conditions under which we observed that oxLDL triggered increased TUNEL, Annexin V (Fig. 6A-J) and cleaved caspase 3 staining [1]. After 8 h of oxLDL treatment average levels of phospho-MLKL tended to be higher in both wt or pdzk1 KO macrophages, however the differences did not reach statistical significance. After 24 h of oxLDL treatment, phospho-MLKL levels were unchanged in wt macrophages and tended to be higher in pdzk1 KO macrophages but, again, the results did not reach statistical significance (Fig. 9). This is consistent with reports that oxLDL induces necroptosis in the context of apoptosis inhibition (e.g. treatment with the pan-caspase inhibitor peptide zVAD-FMK) [5]. This suggests that in our studies and those reported in [1], treatment of macrophages with oxLDL in the absence of other agents (apoptosis inhibitors) led to induction of apoptosis but not necroptosis. We also saw an apparent trend towards increased levels of phospho-MLKL in pdzk1 KO compared to wt macrophages in the 24 h treatment samples that was not apparent in the 8 hr treatment samples, however the results did not reach statistical significance (Fig. 9). Whether PDZK1 affects necroptosis induction under conditions which have been reported by others [5] to induce necroptosis (namely oxLDL treatment in the presence of apoptosis inhibition with zVAD-FMK) remains to be determined. Wt mice were transplanted with BM from control wt (n ¼ 6) or pdzk1 KO donors (n ¼ 9) and allowed to recover for 8 weeks. Mice were injected i.p. with 1 ml of 10% thioglycollate, to recruit macrophages. 72 h later, mice were injected i.p. with tunicamycin (1 mg/kg body weight) (n ¼ 3 for mice with wt BM and n ¼ 4 for mice with pdzk1 KO BM) or an equivalent volume of DMSO as control (n ¼ 3 for mice with wt BM and n ¼ 5 for mice with pdzk1 KO BM) and then euthanized 24 h later. Peritoneal macrophages were collected by peritoneal lavage, and subjected to CD11b, CC3 and PI staining and flow cytometry. Proportions of CD11bþ cells that were positive for CC3 staining were determined. No PIþ cells were detected. Data was analyzed by 1 way ANOVA with Tukey's multiple comparisons test. NS indicates not statistically significant (p ¼ 0.95). ****p o 0.0001. Fig. 5. HDL protects wt but not pdzk1 KO macrophages from tunicamycin-induced apoptosis measured by CC3 immunostaining. Peritoneal macrophages from wt or pdzk1 KO mice were cultured in parallel in lipoprotein deficient medium and treated for 24 h with tunicamycin (TN) (10 μg/ml) or DMSO control in the presence or absence of HDL (50 μg protein/ml) as indicated. Apoptosis was detected by staining for CC3 (green). Nuclei were counterstained with DAPI (blue). Representative images of n ¼ 3 are shown. Scale bars ¼ 25 μm. Quantification is presented in Fig. 3I

Inactivation of PDZK1 in macrophages does not affect bacterial lipopolysaccharide (LPS) mediated induction of markers of M1 macrophage polarization
Peritoneal macrophages from wt or pdzk1 KO mice were treated in culture without or with LPS and the expression of transcripts corresponding to inflammatory markers were determined by RT-PCR (Fig. 10). LPS treatment of both wt and pdzk1 KO macrophages resulted in induction of interleukin (IL)-1β and IL-6 and monocyte chemotactic protein (MCP)-1 gene expression and inactivation of pdzk1 did not alter this level of induction, suggesting that PDZK1 did not affect macrophage polarization towards an M1 phenotype.

Materials
OxLDL (#J5591/BT-910) and HDL (#J64903/BT-914) were from human sources and were purchased from Alfa Aesar (Tewksbury, MA, USA). Tunicamycin and LPS (from E. coli O111:B4) were from Sigma Aldrich Chemical Co (St. Louis, MO, USA). Antibodies and suppliers are listed in Table 2. All other reagents were obtained as described [1]. Fig. 6. HDL protects wt but not pdzk1 KO macrophages from apoptosis induced by oxLDL or UV irradiation. Peritoneal macrophages from wt or pdzk1 KO mice were cultured in parallel in lipoprotein deficient medium and treated for 24 h with or without oxLDL (100 μg protein/ml) in the presence or absence of HDL (50 μg protein/ml) as indicated. Alternatively, cells were exposed to UV irradiation (50 mJ/cm 2 ) and then treated for 24 h without or with HDL (50 μg protein/ml) as indicated. Apoptosis was detected by staining with FITC-annexin V or TUNEL.

Mice
All procedures involving mice were approved by McMaster University's Animal Research Ethics Board in accordance with Canadian Council on Animal Care guidelines. Sources of mice were described in reference [1].

Bone marrow transplantation and evaluation of tunicamycin-induced apoptosis in vivo
10-week old male wt or ldlr -/mice were transplanted with wt or pdzk1 -/bone marrow (BM) as described in the methods section of the accompanying paper [1]. For ldlr KO mice, four weeks after BM transplantation (BMT), atherosclerosis was induced by feeding the mice a high fat diet for 10 weeks as described in the accompanying paper [1]. For wt mice transplanted with either wt or pdzk1 -/-BM, BMT was carried out as described [1]. Eight weeks after BMT, mice were injected intraperitoneally with thioglycollate and, 72 h later, with tunicamycin (1 mg/kg body weight in 150 mM dextrose) as described in the methods section of the accompanying paper [1]. Mice were euthanized 24 h after tunicamycin injection, peritoneal cells were collected and analyzed by flow cytometry by staining for the myeloid marker CD11b, for apoptosis by staining for CC3 and propidium iodide as described in the methods section of the accompanying paper [1].

Blood cell analysis
Blood was collected by cardiac puncture into heparinized tubes. Red blood cell parameters (hematocrit, mean cell volume and distribution width) were determined using a Hemavet Multi-species Hematology System (Drew Scientific, Miami Lakes, FL, USA). For flow cytometry analysis of leukocytes, erythrocytes were lysed by incubating 0.2 mL of blood with 2.0 mL of 1 Â Flow Cytometry Mouse Lysis Buffer (R&D Systems, Minneapolis, MN, USA) for 10 min at room temperature. Afterwards, cells were pelleted (1200 rpm for 5 min in a microfuge at 4°C), washed twice with FACS buffer (PBS containing 1% BSA) and labeled by incubation on ice for 1 hr with the following antibodies diluted 25-fold in FACS buffer: either FITC-labeled rat anti-mouse CD3, or both PerCP-Cy5.5 anti-hu/mo CD45R/B220 and APC anti-mouse CD11b. Flow cytometry was performed using a BD FACScalibur TM flow cytometer (BD Biosciences, San Jose, CA, USA). Data was processed by FlowJo data analysis software (FlowJo, LLC., Ashland, OR, USA).

Monocyte Recruitment
Monocyte recruitment into atherosclerotic plaques was analyzed by labeling circulating monocytes with fluorescent beads, as previously described [7,8]. Ldlr KO mice that had been transplanted with BM from wt or pdzk1 KO donors and then fed the high fat, atherogenic diet for 10 weeks, as described in the accompanying paper [1], were injected i.v. with 250 μl PBS containing 1.5 Â 10 11 Fluoresbrite s YG microspheres (0.5 μm, Polysciences, Inc., Warrington, PA, USA). 24 h after injection, mice were euthanized, and hearts were harvested and frozen in Shandon Cryomatrix (Thermo Fisher Scientific, Ottawa, ON, Canada). 10 μm transverse cryosections of aortic sinus were stained with oil red O. Fluorescence and brightfield images were captured using a Zeiss Axiovert 200 M inverted fluorescence microscope (Carl Zeiss Canada Ltd. Toronto, ON, Canada). The number of green fluorescent beads were quantified as previously described [8].

Immunofluorescence staining for KDEL and Ki67 in atherosclerotic plaques
To determine ER stress in macrophages in atherosclerotic plaques, ER chaperone proteins were detected with a mouse anti-KDEL mAb using Vector s M.O.M.™ immunodetection kit (Vector Laboratories, Inc., Burlingame, CA, USA) with an Alexa-594 streptavidin secondary reagent. Macrophages were stained with rat anti-mouse CD107b (Mac3) antibody followed by Alexa-488 labeled goat anti-rat antibody. Cell proliferation was determined by staining atherosclerotic plaques with rabbit monoclonal (SP6) Ki67 antibody, followed by Alexa-488 labeled goat anti-rabbit secondary antibody. Sections were also costained with DAPI to visualize nuclei. Fluorescent images were captured using a Zeiss Axiovert 200 M inverted fluorescence microscope (Carl Zeiss Canada Ltd. Toronto, ON, Canada).  , treated in culture with 10 ng/ml LPS for 6 h. GAPDH was used as a control and data is presented as mean 7 SEM fold change (n ¼ 4), relative to untreated wt cells. Data was analyzed by 2 way ANOVA with Tukey multiple comparisons test. NS indicates no statistically significant differences: p 4 0.27 for panels A and B and p 4 0.06 for panel C.

Preparation, culture and treatment of peritoneal macrophages
Thioglycollate elicited peritoneal macrophages were prepared from mice as described [1]. Cells (1.5 Â 10 5 /well) were cultured in 8-well Nunc Lab-Tek II Chamber Slides (Thermo Scientific, Waltham, MA, USA) and treated with different agents as described [1]. Agents and concentrations used included: tunicamycin (10 μg/ml); oxLDL (100 μg protein/ml); HDL (50 50 μg protein/ml). Controls contained an equivalent amount of vehicle (0.1% DMSO for tunicamycin or saline for oxLDL). For UV irradiation, cells in chamber slides (with lids removed) were exposed to 50 mJ/cm 2 of UV irradiation using a UVC-508 UV Crosslinker (Ultralum Inc, Clairmont CA, USA). Immediately following UV irradiation, cell culture media was replaced with fresh media containing or lacking HDL at the concentrations indicated and cells were cultured for 24 h prior to apoptosis analysis by TUNEL staining as described [1].