Positron emission tomography and functional characterization of a complete PBR/TSPO knockout

Abstract

The evolutionarily conserved peripheral benzodiazepine receptor (PBR), or 18-kDa translocator protein (TSPO), is thought to be essential for cholesterol transport and steroidogenesis, and thus life. TSPO has been proposed as a biomarker of neuroinflammation and a new drug target in neurological diseases ranging from Alzheimer's disease to anxiety. Here we show that global C57BL/6-Tspo(tm1GuWu(GuwiyangWurra))-knockout mice are viable with normal growth, lifespan, cholesterol transport, blood pregnenolone concentration, protoporphyrin IX metabolism, fertility and behaviour. However, while the activation of microglia after neuronal injury appears to be unimpaired, microglia from (GuwiyangWurra)TSPO knockouts produce significantly less ATP, suggesting reduced metabolic activity. Using the isoquinoline PK11195, the ligand originally used for the pharmacological and structural characterization of the PBR/TSPO, and the imidazopyridines CLINDE and PBR111, we demonstrate the utility of (GuwiyangWurra)TSPO knockouts to provide robust data on drug specificity and selectivity, both in vitro and in vivo, as well as the mechanism of action of putative TSPO-targeting drugs.

Figure 1. Generation and confirmation of global Tspo^−/− mice.

(a) The Tspo gene was knocked out using a targeting construct with loxP sites flanking exons 2 and 3 (also see Supplementary Fig. 1). (b,c) Southern blot analysis (b) and PCR (BV2 mouse microglia were used as the positive control) (c) demonstrated the correct targeting of the Tspo gene. (d,e) Tspo mRNA expression across 13 tissues (triplicates, mean and standard deviation; normalized to Gapdh and Actb) (d) and measurement of TSPO protein by western blot of lysates from the kidney, spleen and testis (e) confirmed the complete absence of any Tspo gene product in Tspo^−/− mice.

Figure 2. Confirmation of global Tspo knockout mice with immunostaining.

(a) Anti-TSPO antibody staining showed the presence of TSPO (here shown in the kidney and testis) in the wild-type and absence in the knockout Tspo^−/− mice. The slides were the same as previously used for autoradiography with the selective TSPO-binding ligand [³H]PK11195 (Fig. 4). (b) Antibody staining of Tspo^+/+ and Tspo^+/− macrophages validates mitochondria as the primary site of the TSPO, which is entirely absent in Tspo^−/− mice. No obvious difference in intracellular density or distribution of the mitochondria was detected in the Tspo^−/− mice (green=TSPO; red=mitochondria; yellow=merged image; blue=nucleus; scale bars: (a) 500 μm and (b) 20 μm).

Figure 3. No constitutive TSPO ligand binding in Tspo^−/− mice.

(a) Tspo^+/+, Tspo^+/− and Tspo^−/− mice are identical in external appearance and general behaviour. However, in vivo imaging (8 males of the same age for each genotype) with PET/CT using the radioligand [¹⁸F]PBR111, the ¹⁸F-labelled analogue to [¹²⁵I]CLINDE, strikingly illustrates that (apart from occasional signals originating from the excretory pathways, such as gut and urinary bladder) Tspo^−/− mice show no ligand binding (thus also demonstrating the selectivity of the used ligand), while both Tspo^+/+ and Tspo^+/− mice have the characteristic distribution of ligand binding, that is, in organs with known high TSPO expression, notably kidney and adrenal gland (green circle). The images are displayed with the colour scaling and are directly comparable (highest values are white). The time-point frame of the PET images is 15–20 min after injection and scaling is 3.8–17.9% ID per cm³. (b) The kinetics of [¹⁸F]PBR111, and the displacement of the radioligand after injection (indicated by arrows) with cold PBR111 (1 mM) to establish non-specific binding, demonstrates that Tspo^−/− mice do not have specific binding of [¹⁸F]PBR111 in any organs, while the ligand kinetic in Tspo^+/+ and Tspo^+/− mice indicates specific binding (ID= injected dose; n=4 for each genotype; error bars denote standard deviation).

Figure 4. Comparative receptor autoradiography and membrane binding.

(a–f) The receptor autoradiographs (16 μm sections) show total binding of [³H]PK11195 (1 nM) and [¹²⁵I]CLINDE (3 nM) as well as competitive binding with 10 μM unlabelled CLINDE (CB(CL)), PK11195 (CB(PK)) and PBR111 (CB(PBR)) in the spleens (a,b), kidneys (c,d) and testes (e,f) of Tspo^+/+, Tspo^+/− and Tspo^−/− mice. Specific binding of 3 nM [¹²⁵I]CLINDE and 1 nM [³H]PK11195 is clearly visible in tissue sections from Tspo^+/+ and Tspo^+/− mice and is displaceable by the unlabelled ligands. There is no specific binding in Tspo^−/− tissue. (g,h) Specific binding using [³H]PK11195 in testicular tissue (g) and kidney tissue (h) (n=3 for each genotype Tspo^+/+, Tspo^+/− and Tspo^−/−). Tspo^+/− mice (kidney (Bmax and Kd): 45,361.0 fmol per mg and 5.94 nM; testes: 3,696.0 fmol per mg and 7.32 nM) have approximately half the binding of Tspo^+/+mice (kidney: 108,934.0 fmol per mg and 12.90 nM; testes: 4,920.0 fmol per mg and 2.66 nM) while Tspo^−/−mice have no appreciable binding. Curves represent non-linear regression of experimentally obtained data points and data are expressed as percentage relative to Tspo^+/+ specific binding. Error bars denote standard deviation.

Figure 5. Whole-body receptor autoradiography of neonatal mice.

Receptor autoradiography using the TSPO ligand [¹²⁵I]CLINDE on whole bodies of 2-day-old neonatal mice. The autoradiographs show total binding of [¹²⁵I]CLINDE (3 nM) as well as competitive displacement binding with 10 μM unlabelled CLINDE (CB(CL)), PK11195 (CB(PK)) and PBR111 (CB(PBR)). Specific binding of [¹²⁵I]CLINDE is clearly visible in the Tspo^+/+ mice and is displaceable by all three unlabelled ligands. Non-specific binding is seen in areas of nuchal fat or stomach content.

Figure 6. No inducible TSPO ligand binding in Tspo^−/− mice.

(a,b) TSPO-binding ligands PK11195 and CLINDE/PBR111. (c) Axotomy of the facial nerve induces a retrograde neuronal reaction and highly reproducible microglial activation in the injured facial nucleus. (d) Autoradiography with [³H]PK11195 and [¹²⁵I]CLINDE and immunohistochemical staining of the microglial activation marker CD11b on consecutive brain sections confirmed the previously reported localized induction of TSPO ligand binding in the injured facial nucleus contemporaneous to the activation of microglia of Tspo^+/+ animals. (e) In contrast, no binding of [³H]PK11195 and [¹²⁵I]CLINDE could be induced in Tspo^−/− mice despite the undiminished presence of activated microglia in the injured facial nucleus, thus providing evidence that the high selectivity of [³H]PK11195 and [¹²⁵I]CLINDE for their respective binding sites on the TSPO is retained in pathologically changed tissue. (f) Immunofluorescent anti-CD11b staining of activated microglia in the injured facial nucleus of Tspo^−/− mice revealed no obvious differences in microglial activation, with its characteristic localization of activated, perineuronal microglia, here (g) shown as higher magnification of the above section. Scale bars in (d,e) and the left image of (f) denote 1 mm, and that in the right image of (f) and in (g) 100 μm. The asterixes in (g and f) indicates the soma of neurons.

Figure 7. PBR/TSPO-expressing brain tumour in global Tspo^−/− mouse brain.

(a) [¹⁸F]PBR111 PET/CT demonstrates a high-contrast signal confined exclusively to the Tspo^+/+ brain tumour, while Tspo^−/− surrounding brain shows no ligand binding. (b,c) For the same animal shown in (a), in vitro autoradiography with [³H]PK11195 (b) and subsequent immunohistochemical staining with anti-TSPO antibody on the same brain sections (c) confirm that PBR/TSPO ligand binding is strictly limited to the brain tumour region and no PBR/TSPO binding or recognition sites are present in Tspo^−/− tissue. Scale bars, 1 mm. The PET images show the time frame from 5 to 10 min after injection and scaling is 0–5.3% ID per cm³ for (a).

Figure 8. Functional characterization of global Tspo^−/− mice.

(a) No significant differences in body length between Tspo^+/+ (n=5) and Tspo^−/− (n=9) neonatal 1-day-old mice are apparent (Student’s t-test). (b) Likewise, there is no significant difference (ANOVA) in the weight gain trajectory measured from 4 to 83 weeks in Tspo^+/+, Tspo^+/− and Tspo^−/− mice within the same sex (n=10–80 animals per time point for weeks 4–43 per group; n=3–44 per time point for weeks 44–83). Independent from genotypes, females weigh significantly less than males, as is known to be the case for this mouse strain (b). (c–e) No significant differences between Tspo^+/+ and Tspo^−/− mice are found in blood pregnenolone concentrations (n=4 for all groups; Student’s t-test) (c); blood protoporphyrin IX (PPIX; left: three typical closely overlapping spectra from the different genotypes after addition of 5-aminolevulinic acid (ALA), right: no significant differences in levels of PPIX (PPIX per ml of blood in arbitrary units (A.U.); n=3–5 per genotype; Student’s t-test)) (d); and mRNA expression of steroidogenic acute regulatory (StAR) protein, P450scc and Tspo2 in nine organs (normalized to reference genes Actb and Gapdh, n=3 per genotype) (e). Error bars denote standard deviation.

Figure 9. Decreased metabolic activity in Tspo^−/− microglia.

(a–c) The basal, that is, mitochondrial and non-mitochondrial, oxygen consumption rate (OCR) is consistently and significantly lower in Tspo^−/− microglia compared to Tspo^+/+ microglia (Tspo^−/−: 49.0±16.3 pmol per min; n=95; Tspo^+/+: 81.0±21.8 pmol per min, n=154; Student’s t-test), while OCR after application of the ATPase inhibitor oligomycin (3 μM) is reduced to similar levels in both genotypes (n=11–14 per group) (a). Thus, the oligomycin-inhibitable mitochondrial ATP production in Tspo^−/− microglia is significantly lower than in Tspo^+/+ microglia (*P≪0.01; Student’s t-test) (c). Basal mitochondrial OCRs in Tspo^−/− and Tspo^+/+ microglia after inhibition of the electron transport chain complexes I and III with 10 μM rotenone+10 μM antimycin A are reduced by similar amounts (c) and to similar levels (b) (n=11–14 per group). (d) Metabolic activity is decreased in Tspo^−/− microglia. Basal OCR vs ECAR (extracellular acidification rate) shows that both OCR and ECAR in Tspo^−/− microglia (n=95) are lower than those in Tspo^+/+ microglia (n=154). Error bars denote standard deviation for (a–d). Further extensive hematological and behavioural data are presented in Supplementary Tables 2 and 3.