3.2.4 Increased expression and activity of P-gp and BCRP
A selective increase in microvascular expression of P-gp and BCRP during disease progression in three ALS mouse models (SOD1G93A, SOD1G86R and TDP43A315T) has been reported in the cerebral cortex and spinal cord (Jablonski et al., 2012; Milane et al., 2010). mRNA levels were also significantly increased in lumbar spinal cord homogenates 2.13-fold for P-gp and 1.72-fold for BCRP in mice at symptomatic stages compared to those at presymptomatic stages (Jablonski et al., 2012). In addition, Western blot was employed to assess P-gp and BCRP abundance in the whole spinal cord and cerebral cortex of SOD1G93A mice. Increased expression of P-gp (1.96-fold) and BCRP (1.69-fold) was observed in the spinal cord of symptomatic mice compared to presymptomatic SOD1G93Amice and elevated levels of P-gp (1.35-fold) and BCRP (1.28-fold) were reported in the cerebral cortex of symptomatic mice compared to presymptomatic SOD1G93A mice (Jablonski et al., 2012). The increased expression of P-gp and BCRP was demonstrated in another mouse model of ALS by Jablonski et al. comparing P-gp and BCRP levels in the spinal cord and cerebral cortex of symptomatic TDP43A315T mice with presymptomatic mice. An independent group (Milane et al., 2010) demonstrated increased P-gp abundance (1.5-fold) in microvessels isolated from the brains of presymptomatic SOD1G86R mice compared to age-matched wildtype controls. However, no alteration of BCRP abundance was detected in these SOD1G86R mice.
In SOD1G93A rats, microvessels from brain and spinal cord sections were assessed immunohistochemically and microvessels were also isolated for protein quantification using Western blot (Chan, Evans, Banks, Mesev, Miller & Cannon, 2017). In this study Chanet al . demonstrated immunohistochemically that P-gp protein expression was significantly increased in the cerebral cortex (1.88-fold) and the spinal cord (1.46-fold) from symptomatic SOD1G93A rats compared to age-matched wildtype rats. In addition, the expression of BCRP was not affected when assessed immunohistochemically. Given immunohistochemistry is semi-quantitative, protein expression in isolated microvessels was reassessed using Western blot, and a 1.5-fold increase in P-gp expression was demonstrated in isolated microvessels from brain and spinal cord of symptomatic SOD1G93A rats. It was also noted that microvascular BCRP expression increased 1.15- and 1.25-fold in brain and spinal cord, respectively, although statistical significance was not achieved. These studies clearly demonstrate increased expression of P-gp and sometimes BCRP in the brain and spinal cord in models of ALS. Some differences in the results were observed. For example, whole brain and spinal cord homogenates were used by Jablonski et al . and the samples obtained from the microvascular isolation techniques employed by Chanet al. and Milane et al. often contain astrocytes, which express P-gp (Golden & Pardridge, 1999).
In addition to assessing expression of P-gp and BCRP, Jablonski et al. assessed their activity using a confocal microscopy-based transport assay where microvessles were incubated with fluorescent substrates (NBD-cyclosporin A for P-gp; bodipy-prazosin for BCRP) (Jablonski et al., 2012). Luminal substrate accumulation was measured in the absence and presence of specific transport inhibitors (PSC833 for P-gp, Ko143 for BCRP). This assay successfully demonstrated that the transport activity of P-gp was increased 1.8-fold and 2-fold in brain and spinal cord microvessels, respectively, from symptomatic SOD1G93A mice relative to age-matched wildtype mice and presymptomatic SOD1G93A mice. In a follow up study, the in vivo accumulation of a specific P-gp substrate, LD800, in the spinal cord of symptomatic SOD1G93A mice was assessed and compared to wildtype mice following intraperitoneal administration (Jablonski et al., 2014). Spinal cord levels of LD800 levels were significantly lower in SOD1G93A mice comparing to wildtype mice, suggesting increased P-gp activity at the BSCB, in line with increased P-gp expression. A similar study was performed, where a P-gp substrate [3H]-digoxin and a BCRP substrate [3H]-prazosin were dosed intraperitoneally to both SOD1G86R mice and wildtype mice (Milane et al., 2010). A reduced disposition of [3H]-digoxin (1.5-fold) but not [3H]-prazosin was observed in the brain of presymptomatic SOD1G86R mice compared to wildtype controls, suggesting increased function of P-gp but not BCRP in this mouse model. This alteration is in line with reduced P-gp protein abundance (1.5-fold) and unchanged BCRP abundance observed in this mouse model.
Chan et al. demonstrated a 2-fold increase in P-gp activity in microvessels isolated from the brain and spinal cord of symptomatic SOD1G93A rats compared to those from presymptomatic rats using the same confocal microscopy-based transport assay mentioned above (Chan, Evans, Banks, Mesev, Miller & Cannon, 2017). Overall, these studies demonstrate increased P-gp and BCRP abundance and activities at the CNS barriers in ALS rodent models, suggesting that the barriers may be more restrictive to substrates of these transporters. This is of particular concern given that riluzole, which is effective in early ALS and less effective in late stage ALS, is a substrate of P-gp and BCRP (Zoccolella et al., 2007). The reduced effectiveness of riluzole in SOD1G93A rats may result from P-gp/BCRP overactivity, as less riluzole may be able to enter the CNS. In fact, a 1.7-fold reduction in riluzole disposition into the brain of SOD1G86R has been reported compared to wildtype mice following intraperitoneal administration (Milane et al., 2010). Therefore, approaches to overcome increased P-gp and BCRP activities in ALS may improve the effectiveness of riluzole. This has been confirmed in SOD1G93A mice that received riluzole with or without elacridar (an inhibitor of P-gp and BCRP) (Jablonski et al., 2014). The efficacy of riluzole improved when co-administered with elacridar and this was associated with increased riluzole delivery to the brain. However, such generic inhibition of P-gp/BCRP comes with side effects, and therefore, an approach to restore P-gp/BCRP to levels similar to the healthy CNS barriers is more likely to be a more effective clinical strategy. In addition, while much has been reported on P-gp and BCRP in ALS, our knowledge on other transporters at the CNS barriers is limited. Profiling the expression at the CNS barriers could potentially identify efflux transporters that should be avoided for effective CNS delivery or influx transporters, which could be effectively targeted to increase CNS exposure of drugs intended to reach the CNS.