Challenges in pericyte research: pericyte-specific and subtype-specific markers

Abstract

Pericytes are a heterogenous population that plays multiple important roles in both physiological and pathological conditions. Although many markers and transgenic mouse lines have been used to identify pericytes, these tools all have limitations. For example, many of them are not pericyte-specific and none of them are able to distinguish different subtypes of pericyte. Here, we summarize commonly used pericyte markers and transgenic mouse lines, compare their unique features and limitations, and discuss key points to consider when using these tools or interpreting data generated by using them. Identifying/developing pericyte-specific and subtype-specific markers/tools will fill the gap of knowledge and substantially move the field forward.

Introduction

Peircytes are perivascular cells that cover small blood vessels. Recent studies have demonstrated that pericytes play important roles in angiogenesis, blood-brain barrier integrity maintenance and cerebral blood flow regulation [1]. In addition, pericytes also display stem cell properties and are able to differentiate into various other cell types, such as fibroblast-like cells and microglia/macrophages, in pathological conditions [1]. Consistent with these important functions, pericytes are actively involved in the pathogenesis of Alzheimer’s disease and stroke, and thus may be targeted in the treatment of these neurological disorders. It should be noted, however, that there are two major challenges in pericyte research, which may be responsible for some controversial findings in the field.

First, there are no pericyte-specific markers available currently. Many markers have been used to identify pericytes, such as platelet-derived growth factor receptor beta (PDGFRβ), aminopeptidase N (APN or CD13), regulator of G protein signaling 5 (RGS5), and chondroitin sulfate proteoglycan 4 (CSPG4). Correspondingly, several transgenic mouse lines, including the PDGFRβ-EGFP (Gensat.org. Strain name: Tg(Pdgfrb-EGFP)JN169Gsat), PDGFRβ-RFP [2], PDGFRβ-Thymidine Kinase (TK) [2], PDGFRβ-Cre [3], PDGFRβ-CreER^T2 [4], RGS5-EGFP [5], CSPG4-YFP [2], CSPG4-dsRed [6], CSPG4-TK [2], CSPG4-Cre [7], CSPG4-CreER^T2 [8], have been generated to target pericytes. None of these markers, however, are pericyte-specific. For example, PDGFRβ, CD13 and RGS5 also mark vascular smooth muscle cells (vSMCs) and fibroblasts in addition to pericytes, while CSPG4 labels pericytes, vSMCs and oligodendrocyte precursor cells. Therefore, the “pericyte” population identified using these markers/tools most likely contains contaminating cells, which makes it extremely challenging to target pericytes specifically for in vivo loss-of-function studies (e.g. pericyte-specific conditional knockout) and isolate high-purity live pericytes for in vitro studies.

To address this issue, an innovative dual-promoter approach (PDGFRβ-Flp:CSPG4-FSF-CreER), in which Cre expression is driven by both PDGFRβ and CSPG4 promoters, has been developed [9]. Specifically, the PDGFRβ prmoter drives expression of Flp recombinase, which removes the Frt-flanked Stop sequence, enabling epression of CreER under the CSPG4 promoter. It has been reported that pericytes are exclusively labeled in PDGFRβ-Flp:CSPG4-FSF-CreER mice after tamoxifen injection [9], highlighting the specificity of the dual-promoter system. However, it should be noted that this approach involves two (PDGFRβ and CSPG4) promoters and two (Flp and Cre) recombinases. This sophisticated genetic design requires longer time for mouse breeding, which limits its use in the field. In addition, a recent study identified Atp13a5, a P-type ATPase and a cation transporter, as a CNS pericyte-specific marker [10]. Specifically, using the Atp13a5–2A-CreERT2-IRES-tdTomato knock-in mice, the authors found tdTomato expression in pericytes but not other cells in the CNS. More importantly, tdTomato was exclusively detected in CNS (brain, spinal cord and retinal) pericytes, but not pericytes from peripheral organs (e.g. heart, liver and kidney). These findings suggest that Atp13a5 may be a CNS pericyte-specific marker under homeostatic conditions. It remains unclear whether Atp13a5 specifically and reliably labels CNS pericytes in pathological conditions, such as stroke and Alzheimer’s disease. This important question needs to be answered in future research.

Second, there are no subtype-specific markers for pericytes. As a heterogenous population with multiple developmental origins [1], pericytes demonstrate distinct features in different organs (e.g. brain vs. muscle) or even within the same organs (e.g. ensheathing vs. capillary pericytes in the brain). Unfortunately, none of the above-mentioned markers and genetic tools are able to distinguish different subtypes of pericytes. For example, all pericyte populations are labeled in the PDGFRβ/CSPG4 dual-promoter system [9]. Although Atp13a5 marks CNS pericytes exclusively, it is unable to discern type I (with long processes covering microvessels) and type II (with short processes wrapping around the whole microvessels) pericytes in the CNS [10]. The lack of markers to label different subtypes of pericytes make it impossible to investige pericyte biology/function in a subtype-specific manner. This is a critical barrier to progress in the field. Identifying subtype-specific pericyte markers and developing genetic tools that enable targeting pericyte subtypes will fill the gap of knowledge and substantially move the field forward.

The features and limitations of commonly used pericyte markers/tools are summarized in Table 1. What can we do with these imperfect markers and tools? First, more rigorous experimental design should be used to overcome the limitations of these markers/tools. In case that pericyte markers also label other cell types, the contaminating cells should be controlled. For example, when PDGFRβ-CreER^T2 is used to ablate target gene in pericytes, vSMC-specific (SM22α-Cre or Myh11-CreER^T2) and fibroblast-specific (Col1α1-Cre or Col1α2-CreER^T2) Cre lines should be included as controls. If a phenotype is observed in the PDGFRβ-CreER^T2 line but not vSMC-specific or fibroblast-specific mutants, it is more likely caused by loss of target gene in pericytes. Next, care should be taken when interpreting results obtained from using these imperfect markers/tools, since other cells (non-pericytes) may be targeted and multiple subtypes of pericytes are targeted. For instance, if the sophisticated dual-promoter system is used to target pericytes specifically and a brain phenotype (e.g. blood-brain barrier disruption) is observed, this phenotype may be caused directly by defects in brain pericytes or indirectly by changes secondary to deficits in peripheral pericytes. Third, major findings in mice should be validated in vitro and/or ex vivo. The relatively simple systems (e.g. in culture) allow direct investigation of target cell biology. Consistent pericyte changes in in vitro and in vivo experiments will strengthen the conclusions.

Table 1.

Summary of commonly used pericyte markers and genetic tools

Markers	Target Cells	Tools	References
PDGFRβ	Pericytes, vSMCs, Fibroblasts	PDGFRβ-EGFP	Gensat.org. Strain name: Tg(Pdgfrb-EGFP)JN169Gsat
		PDGFRβ-RFP	[2]
		PDGFRβ-TK	[2]
		PDGFRβ-Cre	[3]
		PDGFRβ-CreERT2	[4]
CD13	Pericytes, vSMCs, Fibroblasts	-	-
RGS5	Pericytes, vSMCs, Fibroblasts	Rgs5-EGFP	[5]
CSPG4	Pericytes, vSMCs, Oligodendrocyte precursor cells	CSPG4-YFP	[2]
		CSPG4-dsRed	[6]
		CSPG4-TK	[2]
		CSPG4-Cre	[7]
		CSPG4-CreERT2	[8]
PDGFRβ & CSPG4	Pericytes	PDGFRβ-Flp:CSPG4-FSF-CreER	[9]
Atp13a5	CNS pericytes	Atp13a5–2A-CreERT2-IRES-tdTomato	[10]