Splenectomy induces biochemical remission and regeneration in experimental murine autoimmune hepatitis

The animal care and experiments were performed in accordance with the institutional and national guidelines. All animal experiments were performed according to protocols approved by the animal welfare commission of the Hannover Medical School and local ethics animal review board (Lower Saxony State Office for Consumer Protection and Food Safety, Oldenburg, Germany).

NOD/Ltj mice were bred and maintained under specific pathogen-free conditions at the central animal facility of the Hannover Medical School (Hannover, Germany). Mice were injected intravenously with a total of 4 × 10⁹ infectious particles containing an adenovirus expressing formiminotransferase cyclodeaminase (Ad-FTCD) in PBS. Mice (n = 5) were splenectomized 12 weeks later in the late stage of disease [9]. All mice (n = 10) were sacrificed 18 weeks after infection, and then all analyses were performed.

Ad-FTCD was generated as described previously [9]. In brief, FTCD was amplified from human liver cells. The constructs were fused to the Ad transfer vector pShuttle-CMV (Stratagene). By homologous recombination, this shuttle vector was recombined with pAdEasy-1, which carries deletions in the E1 and E3 regions (Stratagene). The generated adenovirus genome can be amplified only within the HEK 293 packaging cell line, complementing the essential regions. Purification of recombinant adenovirus was performed using a cesium chloride gradient, and the adenoviral stocks were quantified using an Adeno-X™ Rapid Titer Kit (Clontech).

To obtain single-cell suspension, livers were minced by a gentleMACS Dissociator (Miltenyi Biotec, Bergisch Gladbach). Then, red blood cells were lysed (eBioscience™ 1X RBC Lysis Buffer, Thermo Fisher Scientific) and IHLs (intrahepatic lymphocytes) were separated using a 40%/70% Percoll (GE Healthcare) gradient. Subsequently, lymphocytes were stained with anti-CD3, anti-CD4, anti-CD8, anti-Ki-67, anti-B220, anti-Foxp3, and anti-CD62L antibodies. All image acquisitions were performed with an LSRII SORP interfaced with DIVA software (BD Biosciences) as described previously [16, 19].

Blood samples were collected via the retroorbital route before sacrificing the mice. Aspartate aminotransferase (AST) and alanine transaminase (ALT) levels were determined by photometric enzyme activity assays with an Olympus AU400 Chemistry Analyzer using serum as described previously [9, 20].

Murine liver tissue was fixed in formalin and embedded in paraffin or only embedded in Tissue-Tek® O.C.T.™ Compound (Sakura) for cryosections (8 µm). Paraffin-embedded sections (5 µm) were prepared for hematoxylin and eosin (HE) staining and analyzed with an AxioImagerM1 using AxioVision 4.8 software (Zeiss). The sections were further examined in a blinded manner by a pathologist using the approved modified hepatitis activity index (mHAI) for autoimmune hepatitis as described previously [5, 21]. Immunofluorescence microscopy was performed as previously described [11‐13]. Briefly, 4 µm cryosections were fixed with acetone, rehydrated, blocked, stained with anti-CD4, anti-CD8, anti-Foxp3 and DAPI and analyzed with an AxioImagerM1 as above.

Proteins were measured using an Olink® MOUSE EXPLORATORY panel* (Olink Proteomics AB, Uppsala, Sweden) according to the manufacturer’s instructions as described previously [1, 22]. The proximity extension assay (PEA) technology used for the Olink protocol has been well described [23] and allows for 92 analytes to be analyzed simultaneously. Briefly, pairs of oligonucleotide-labeled antibody probes bind to their targeted protein, and if the two probes are brought in close proximity, the oligonucleotides will hybridize in a pairwise manner. The addition of DNA polymerase leads to a proximity-dependent DNA polymerization event, which generates a unique PCR target sequence. The resulting DNA sequence is subsequently detected and quantified using a microfluidic real-time PCR instrument (Biomark HD, Fluidigm). The data are then quality-controlled and normalized using an internal extension control and an interplate control to adjust for intra- and inter-run variations. The final assay read-out is presented in Normalized Protein eXpression (NPX) values, which is an arbitrary unit on a log2-scale in which a high value corresponds to high protein expression. All assay validation data (detection limits, intra- and interassay precision data, etc.) are available on the manufacturer’s website (www.​olink.​com).

RNA was isolated, cDNA was preamplified, and quantitative RT-PCR was performed as described previously [2, 19]. Normalization of the Ct was performed by subtracting the mean values of the housekeeping genes glyceraldehyde-3-phosphate dehydrogenase (Gapdh) and actin beta (Actb) from those of genes of interest. Heatmaps and PCA plots of the − delta Ct values were created via Qlucore software (p < 0.05 and q < 0.2).

Unpaired Student 2-tailed t tests, correlation analysis, principal component analysis (PCA) and heatmap analysis were performed using GraphPad Prism version 7.00 for Mac and Qlucore Omics Explorer 3.5. * significant difference with p ≤ 0.05; ** very significant difference; p ≤ 0.01; p > 0.05 was considered to be nonsignificant (ns).

The pathogenesis and pathophysiology of AIH are still largely unknown. Likewise, the roles of different lymphocyte populations are still under discussion. Therefore, we induced emAIH with an adenovirus (Ad) encoding FTCD (Ad-FTCD) and removed the spleen at week twelve, which was during the late stage of the disease [9]. After 6 weeks, the mice were sacrificed (Fig. 1A). Serum samples were collected and analyzed, and we quantified 92 different proteins within them with Olink technology (Fig. 1B). In total, 9 of the proteins were differentially regulated with clustering within the two groups (Fig. 1C), with three exhibiting significant changes (p < 0.001) (Fig. 1D). Protein analyses demonstrated the downregulation of the inflammatory proteins chemokine (C–C motif) ligand 3 (CCL3), interleukin (IL)-1a and cadherin (Cdh)6 as well as Axin1 and quinoid dihydropteridine reductase (Qdpr). On the other hand, proteins related to regeneration such as Matrilin (Matn)-2, Glial cell-derived neurotrophic factor (Gdnf) and the myokine IL‐6 were upregulated. Notably, the chemokine Cxcl9 was upregulated and thought to be an inflammatory chemokine.

We also analyzed TH1-, TH2-, TH17- and fibrosis-related markers in liver tissue by quantitative PCR, but we could not find any differentially regulated genes in a set of 21 genes (data not shown).

In conclusion, splenectomy during the late stage of a chronic inflammatory disease shaped regeneration at the molecular level.

We previously showed that splenectomy does not affect hepatic histology [1, 2]. Nonetheless, AIH is a chronic and progressive disease. Thus, the pathology at the histological and biochemical levels after 18 weeks was completely unknown with respect to the observed regenerative molecular pattern.

Therefore, liver sections were produced and analyzed by microscopy. No obvious differences in histology could be observed at 6 weeks after splenectomy between splenectomized and non-splenectomized mice with emAIH (Fig. 2A). The same was true for the mHAI, which was higher than in comparison to the scores 6 weeks earlier from other studies [5, 9] but comparable between the two groups (Fig. 2B). Additionally, the average size of lymphatic infiltrates was just slightly smaller (Fig. 2C), while the liver weight was unchanged (Fig. 2D). However, biochemical analysis of the serum revealed a consequence of the regenerative molecular pattern observed. While the levels of aspartate transaminases (ASTs) were not prominently altered (Fig. 2E), the levels of alanine transaminases (ALTs) were significantly reduced by approximately 40% (Fig. 2F).

In conclusion, splenectomy induced biochemical remission of AIH.

The regenerative molecular pattern and biochemical remission of AIH after splenectomy were revealed, but the cellular mechanism remained unknown.

Therefore, we analyzed the cellular composition of IHLs. In contrast to the changes observed in other models, the removal of the spleen, which resulted in the loss of a lymphocyte site, was compensated by an increase in lymphocytes in the liver (Fig. 3A). Here, the relative numbers of B cells were increased after splenectomy (Fig. 3B). Consequently, the relative number of T cells decreased (Fig. 3C). The IgG serum level did not subsequently change (Fig. 3D).

Given that the spleen was removed, the increase in B cells (Fig. 3E) and T cells in absolute numbers was not surprising as the liver is a harbor for lymphocytes (Fig. 3F). To investigate the outcome of the regenerative serum protein pattern and cause of biochemical remission, other cells were evaluated. Analyses revealed that both the relative and absolute numbers of intrahepatic Tregs were increased (Fig. 3G). This supraproportional increase even within the enlarged T-cell compartment of the immune regulatory and tissue-repairing cell subpopulation was meaningful. Consequently, the number of activated CD8⁺ T cells was decreased (Fig. 3H). In summary, the most prominent immunoregulatory cell population was supraproportionally increased in the liver.

In addition to all intrahepatic lymphocytes, we analyzed the local population within the portal inflammation. For this purpose, we prepared cryosections from liver tissue. To reduce the immense background, the organs were treated with sucrose before embedding. Frozen organs were sectioned to 4 µm and stained for CD4, CD8 and Foxp3 (Fig. 4A). CD4 (blue) T cells with Foxp3 (red) within the nuclei are easily visible, which are Tregs. In the animals with emAIH that were additionally splenectomized, there was an even more significant increase in Tregs than in the flow cytometric analysis of all IHLs (Fig. 4B).

Given that the CD4/CD8 ration is decreased in other liver diseases if the disease is worsening, we analyzed this in our splenectomy emAIH model.

As in other models and diseases, the ratio of CD4⁺ to CD8⁺ T cells was slightly increased within the shrunken T-cell compartment (Figs. 4C and 5A). The same was true for the ratio of Tregs to Teffs (Fig. 5B). We also noted a slight increase in the Treg/Teff ratio, but this increase was not significant due to the normal error distribution. In contrast, the correlation analysis of Tregs and ALT showed a good correlation between an increase in Tregs and a decrease in ALT (Fig. 5C).

The observed increase in intrahepatic Tregs might have been the cause of the observed biochemical remission.