Ectopic expression of CD74 in Ikkβ-deleted mouse hepatocytes

Abstract

CD74, a Type II membrane glycoprotein and MHC class II chaperone involved in antigen processing, is normally expressed by cells associated with the immune system. CD74 also forms heterodimers with CD44 to generate receptors to macrophage migration inhibitory factor (MIF), a proinflammatory cytokine. Following targeted Alb-Cre-mediated deletion of Ikkβ in Ikkβ^Δhep mice (Ikkβ^F/F:Alb-Cre, a strain highly susceptible to chemically-induced hepatotoxicity and hepatocarcinogenesis), CD74 is expressed abundantly by adult hepatocytes throughout liver acini, albeit more intensely in midzonal-to-centrilobular regions. By comparison, CD74 expression is not observed in Ikkβ^F/F hepatocytes, nor is it augmented in the livers of Ikkβ^+/+:Alb-Cre mice; CD74 is barely detectable in cultured embryonic fibroblasts from Ikkβ^-/- mice. Microarray profiling shows that constitutive CD74 expression in Ikkβ^Δhep hepatocytes is accompanied by significantly augmented expression of CD44 and key genes associated with antigen processing and host defense, including MHC class II I-Aα, I-Aβ, and I-Eβ chains, CIITA and CD86. Taken together, these observations suggest that Ikkβ^Δhep hepatocytes might express functional capacities for class II-restricted antigen presentation and heightened responsiveness to MIF-signaling, and also suggest further roles for intrahepatocellular IKKβ in the suppression or inactivation of molecules normally associated with the formation and differentiation of cells of the immune system.

Introduction

CD74 is recognized as the invariant chain chaperone (Ii) of MHC class II processing (Bertolino and Rabourdin-Combe, 1996). Its molecular sizes range between 31-45 kDa; the most abundant mouse isoforms, splice variants p31 and p41, are post-translationally modified with chondroitin sulfate side-chains to generate Type II membrane glycoproteins with 30 aa cytoplasmic tails (Bertolino and Rabourdin-Combe, 1996; Stumptner-Cuvelette and Benaroch, 2002). CD74 is expressed throughout the immune system by B cells, activated T cells, dendritic cells, monocytes and macrophages (Bertolino and Rabourdin-Combe, 1996; Calandra and Roger, 2003; Faure-André et al., 2008; Leng and Bucala, 2006; Lue et al., 2006; Stumptner-Cuvelette and Benaroch, 2002). In the liver, CD74 is expressed constitutively in dendritic and Kupffer cells (Momburg et al., 1986), and in resident hepatic stellate cells (HSCs) (Maubach et al., 2007); however, except for one isolated report of a scattered sub-population of CD74-positive hepatocyte clusters (containing 1 to 5 cells) in ♀ B10.BR H-2^k mice (Momburg et al., 1986), CD74 is not expressed in hepatocytes. CD74 is also a component of heterodimeric CD74/CD44 cell surface receptors (Leng and Bucala, 2006; Lou et al., 2006) which mediate mitogenic and proinflammatory effects of macrophage migration inhibitory factor (MIF) (Calandra and Roger, 2003; Leng and Bucala, 2006; Lue et al., 2006), a 12.5 kDa cytokine secreted by a variety of T and B cells, epithelial cells, and liver Kupffer cells (Calandra and Roger, 2003; Kobayashi et al., 1999; Lan, 2008; Leng and Bucala, 2006; Lue et al., 2006). MIF exhibits intrinsic enzymatic activities; induces cyclin D1 expression, mitogenesis and NO production; and plays major roles in the pathogenesis of inflammatory disease (Javeed et al., 2008; Leng and Bucala, 2006; Swant et al., 2005). MIF expression is causally related to chronic hepatitis B virus (HBV) infection (Zhang et al., 2005), fulminant alcoholic hepatitis (Kumagi et al., 2001) and Bacille-Calmette-Guerin-primed lipopolysaccharide (LPS)-induced T cell-mediated liver failure (BLTLF) (Kobayashi et al., 1999). Anti-MIF antisera block BLTLF (Kobayashi et al., 1999) and attenuate experimentally-induced murine colitis (de Jong et al., 2001); anti-sense MIF cDNA attenuates LPS-induced liver injury (Iwaki et al., 2003); and MIF knockout mice are protected from concanavalin A (ConA), LPS- and acetaminophen-induced liver injuries (Bourdi et al., 2002; Bozza et al., 1999; Nakajima et al., 2006). CD74 also is cleaved by regulated intracellular proteolysis (RIP) to form a 10-kDa cytoplasmic polypeptide fragment which migrates into the nucleus and activates nuclear factor kappa B (NF-κB) expression (Becker-Herman et al., 2005). However, the relationship between CD74 and NF-κB in liver disease remains unclear.

To understand how NF-κB, a transcription factor that plays anti-inflammatory and protective roles in hepatotoxic injury (Heyninck et al., 2003) and mediates liver disease, we have investigated an hepatocyte-targeted mouse strain (Ikkβ^Δhep) with Cre recombinase-mediated deletions of IκB kinase β (Ikkβ) exon 3 (Maeda et al., 2003). IKKβ, a subunit of the IκB Kinase (IKK) complex, is required for activation of NF-κB (Ghosh and Karin, 2002; Heyninck et al., 2003). Ikkβ^Δhep mice exhibit defective activation of NF-κB and survival genes, and enhanced susceptibility to hepatotoxins (Maeda et al., 2003, 2005). Following ConA- or LPS+Galactosamine-induced inflammation, Ikkβ^Δhep mice exhibit heightened susceptibility to liver failure, resulting partly from tumor necrosis factor (TNF)α-activation of TNF receptor 2 (TNFR2; Maeda et al., 2003), and from increased reactive oxygen species (ROS) formation. ROS inhibits mitogen activated protein kinase (MAPK) phosphatases and sustains c-JUN nuclear kinase 1 (JNK1) activation (Kamata et al., 2005; Maeda et al., 2003). Following treatment with diethylnitrosamine (DEN), Ikkβ^Δhep mice exhibit increased susceptibility to hepatocellular carcinoma (HCC), resulting partly from cytokine-driven NF-κB activation-independent compensatory hepatocyte proliferation (Laurent et al., 2005; Maeda et al., 2005); from sustained JNK1 activation, accompanied by necrotic hepatocellular interleukin (IL)-1α release which activates IL-1 receptors in Kupffer cells and IL-6 secretion (Sakurai et al., 2008); and, as described most recently, from intrinsic Ikkβ^Δhep hepatocellular growth advantages (Koch et al., 2009).

Neither CD74 nor MIF has been linked to the pathophysiological responses of Ikkβ^Δhep hepatocytes, or to Ikkβ-deletion in other cells in other IKKβ-deficient mice (Chen et al., 2006; Malato et al., 2008). However, CD74 is required for major histocompatibility (MHC) class II antigen processing (Bertolino and Rabourdin-Combe, 1996); and MIF has wide-ranging effects on hepatic inflammation, infection and tumor progression (Bourdi et al., 2002; Bozza et al., 1999; Calandra and Roger, 2003; de Jong et al., 2001; Iwaki et al., 2003; Nakajima et al., 2006; Zhang et al., 2005). In addition, because the proinflammatory effects elicited by the CD74/CD44 MIF receptor, like those by TNFR2 (Maeda et al., 2003), require sustained kinase activation (Leng and Bucala, 2006), we investigated constitutive CD74 expression in Ikkβ^Δhep and Ikkβ^F/F mice.

Surprisingly, we find that Ikkβ^Δhep hepatocytes express CD74, whereas Ikkβ^F/F hepatocytes do not. In addition, as assessed by microarray profiling, the hepatocellular CD74-positive phenotype is accompanied by the expression of large numbers of genes normally associated with immune system processing and host defense mechanisms, including MHC class II I-Aα, I-Aβ, and I-Eβ chains, CD44, the transcriptional coactivator of MHC class II expression CIITA (Ting and Trowsdale, 2002), and coactivator CD86. These observations suggest that targeted Ikkβ deletion confers several unexpected and unusual immunologic phenotypes in hepatocytes which, if functionally active, might have significant implications for both the pathophysiology of liver disease and the regulation of animal cell differentiation.

Materials and Methods

Animals and cell culture

C57BL/6, Ikkβ^+/+:Alb-Cre, Ikkβ^F/F and Ikkβ^F/F:Alb-Cre (referred to as Ikkβ^Δhep) mice were bred at UCSD (La Jolla, CA), maintained and sacrificed according to NIH guidelines (Koch et al., 2009; Maeda et al., 2003). Protocols used in all animal studies were approved by the UCSD IACUC animal ethics committee. Mouse embryonic fibroblasts (MEFs) from Ikkβ^-/- embryos were cultured under standard conditions (Chen et al., 2006).

Isolation and analyses of primary hepatocytes, tissues and macromolecules

Freshly isolated hepatocytes and liver tissues were harvested from adult mice; DNA, RNA and proteins were isolated from hepatocytes, liver tissues and MEFs as described elsewhere (Chen et al, 2006; Koch et al., 2009; Maeda et al., 2003). Ikkβ deletion and genotyping tests were performed by polymerase chain reaction (PCR) using DNA from tails, livers and cultured hepatocytes and MEFs (Chen et al, 2006; Koch et al., 2009; Maeda et al., 2003). RNA and DNA were sized by gel-electrophoresis (Koch et al., 2009).

Microarray profiling

Animals were sacrificed and tissues harvested at 5-6 weeks of age. RNA samples were labeled and hybridized by standard procedures (Feng et al., 2007) using two Illumina WG-6 V2 mouse bead-chips (San Diego, CA, cat. #BD 201-0202) in the UCSD BIOGEM Core (http://microarrays.ucsd.edu/). Each chip carried 6 microarrays housing 47,000 oligonucleotide targets representing ∼5,000 genes/array (see: Fig. 1). Data were normalized and grouped (using the gene ontology databases from the Gene Ontology Consortium [www.genego.org] including pathways from KEGG, Biocarta, and Proteinlounge databases), and analyzed for statistical significance (Feng et al., 2007). Quantitative variation among data samples for individual genes was <1-2%. The original data (see Tables 1 and 2) have been entered into the NCBI GEO Repository (http://www.ncbi.nlm.nih.gov/projects/geo/query/acc.cgi?acc=GSE15476); a sample of selected data in Table 1 has been presented elsewhere (Koch and Leffert, 2010). For GenBank details and probes employed for these analyses, see the Illumina Reference Chip file, MouseWG-6_V2_0_R2_11278593_A (http://www.switchtoi.com/annotationfiles.ilmn).

Figure 1.

Analysis of RNA and DNA from male Cre(-) Ikkβ^F/F and Cre(+) Ikkβ^Δhep mouse livers and freshly isolated hepatocytes used for microarray profiling. (A) EtBr-stained RNA-sizing gel. (B) Deletion analysis. EtBr-stained DNA-sizing gel. Genotyping for deletion confirmation was performed by PCR using standard primers (Koch et al., 2009). M, size markers (bps); lanes #1-#4 as in panel (A); Cre(-) ∼2-kb, Cre(+) ∼0.3-kb. (C) Hybridization format: chips 1 (A) and 2 (B). Each RNA sample (liver tissue [LT], or isolated hepatocytes [H]) was obtained from a separate mouse: Samples #1 and #2, Cre(-)- and Cre(+)-tissues; #3 and #4, Cre(-)- and Cre(+)-isolated hepatocytes. Each sample was hybridized 3×, on 3 separate sectors across the 2 chips, as diagrammed.

Table 1.

Partial results of microarray profiling: Freshly Isolated Liver Tissues.^*

IMMUNE SYSTEM PROCESS p = 6.46685868E-28
Rank	Locus	GenBank Acc.#	Description	E**
2	14961	NM_207105.1	Histocompatibility 2, class II antigen A, beta 1 (H2-Ab1)	-8.86
3	14969	NM_010382.1	Histocompatibility 2, class II antigen E beta (H2-Eb1)	-8.65
4	16149	NM_010545.2	CD74 antigen (invariant polypeptide of major histocompatibility complex)	-8.37
5	14960	NM_010378.2	Histocompatibility 2, class II antigen A, alpha (H2-Aa)	-6.21
6	14963	NM_008199.1	Histocompatibility 2, blastocyst (H2-Bl)	+5.72
8	14998	NM_010386.2	Histocompatibility 2, class II, locus DMa (H2-DMa)	-5.11

DEFENSE RESPONSE p = 2.07679921E-27
Rank	Locus	GenBank Acc.#	Description	E**
4	16149	NM_010545.2	CD74 antigen (invariant polypeptide of major histocompatibility complex)	-8.37
6	14963	NM_008199.1	Histocompatibility 2, blastocyst (H2-Bl)	+5.72
10	17329	NM_008599	Chemokine (C-X-C motif) ligand 9 (Cxcl9)	-4.97
18	15019	NM_023124	Histocompatibility 2, Q region locus 8 (H2-Q8)	-3.82
20	20210	NM_011315	Serum amyloid A 3 (Saa3)	-3.74
29	55985	NM_018866.1	Chemokine (C-X-C motif) ligand 13 (Cxcl13)	-3.31
30	20304	NM_013653.1	Chemokine (C-C motif) ligand 5 (Ccl5)	-3.29
34	12262	NM_007574.1	Complement component 1, q subcomponent, C chain (C1qc)	-3.08
37	17105	NM_017372	Lysozyme (Lyzs)	-2.9
42	12260	NM_009777.1	Complement component 1, q subcomponent, beta polypeptide (C1qb)	-2.85
43	17069	NM_008529.2	Lymphocyte antigen 6 complex, locus E (Ly6e)	-2.82
53	14127	NM_010185.2	Fc receptor, IgE, high affinity I, gamma polypeptide (Fcer1g)	-2.55
67	24088	NM_011905.2	Toll-like receptor 2 (Tlr2)	-2.31
72	21356	NM_009318.1	TAP binding protein (Tapbp)	-2.29
74	15945	NM_021274.1	Chemokine (C-X-C motif) ligand 10 (Cxcl10)	-2.29
78	17110	NM_013590.2	Lysozyme (Lyzs)	-2.25
82	12505	AK045226	CD44 antigen (Cd44)	-2.22

^*Selected expression comparisons between markers from Ikkβ^F/F and Ikkβ^Δhep freshly isolated liver tissues are listed; the markers with these comparatively most disparate expression levels are clustered within two functional groups.

^**Enrichment, E**, as measured by relative microarray chip hybridization signals, is defined as follows: a negative (−) E value signifies that Ikkβ^Δhep > Ikkβ^F/F, and a positive (+) E value signifies that Ikkβ^F/F > Ikkβ^Δhep (because the initial comparative chip data were expressed as X-fold enrichment of hybridization markers from Ikkβ^F/F compared to those from Ikkβ^Δhep RNA samples).

Table 2.

Partial results of microarray profiling: Freshly Isolated Hepatocytes.^*

DEFENSE RESPONSE p = 1.38089039E-21
Rank	Locus	GenBank Acc.#	Description	E**
1	16149	NM_010545.2	CD74 antigen (invariant polypeptide of major histocompatibility complex, class II antigen-associated) (Cd74)	-38.31
19	15953	NM_008330.1	Interferon gamma inducible protein 47 (Ifi47)	-7.15
23	15944	NM_008326.1	Immunity-related GTPase family, M (Irgm)	-6.51
24	17329	NM_008599	Chemokine (C-X-C motif) ligand 9 (Cxcl9)	-6.51
27	93671	NM_053094.1	CD163 antigen (Cd163)	+5.7
34	21356	NM_001025313.1	TAP binding protein (Tapbp)	-5.23
45	17105	NM_017372	Lysozyme (Lyzs)	-4.5
53	12505	AK045226	CD44 antigen (Cd44)	-4.25
55	15024	NM_010395.5	Histocompatibility 2, T region locus 10 (H2-T10)	-4.17
64	21354	NM_013683.1	Transporter 1, ATP-binding cassette, sub-family B (MDR/TAP) (Tap1)	-3.96
65	24088	NM_011905.2	Toll-like receptor 2 (Tlr2)	-3.96
66	12260	NM_009777.1	Complement component 1, q subcomponent, beta polypeptide (C1qb)	-3.96
67	12774	NM_009917.2	Chemokine (C-C motif) receptor 5 (Ccr5)	-3.92
69	17110	NM_013590.2	Lysozyme (Lyzs)	-3.87

IMMUNE SYSTEM PROCESS p = 1.17768030E-18
Rank	Locus	GenBank Acc.#	Description	E**
1	16149	NM_010545.2	CD74 antigen (invariant polypeptide of major histocompatibility complex, class II antigen-associated) (Cd74)	-38.31
3	14961	NM_207105.1	Histocompatibility 2, class II antigen A, beta 1 (H2-Ab1)	-22.48
4	14969	NM_010382.1	Histocompatibility 2, class II antigen E beta (H2-Eb1)	-18.24
6	14998	NM_010386	Histocompatibility 2, class II, locus DMa (H2-DMa)	-15.82
7	14960	NM_010378.2	Histocompatibility 2, class II antigen A, alpha (H2-Aa)	-14.39
8	21822	NM_011579.2	T-cell specific GTPase (Tgtp)	-13.26
9	14468	NM_010259.1	Guanylate nucleotide binding protein 1 (Gbp1)	-11.94
11	14999	NM_010387.2	Histocompatibility 2, class II, locus Mb1 (H2-DMb1)	-10.6
13	15000	NM_010388	Histocompatibility 2, class II, locus Mb2 (H2-DMb2)	-9.36
14	55932	NM_018734.2	Guanylate nucleotide binding protein 3 (Gbp3)	-8.96
20	14469	NM_010260.1	Guanylate nucleotide binding protein 2 (Gbp2)	-7.02
24	17329	NM_008599	Chemokine (C-X-C motif) ligand 9 (Cxcl9)	-6.51
30	60533	NM_021893.2	CD274 antigen (Cd274)	-5.54
33	16912	NM_013585.1	Proteasome (prosome, macropain) subunit, beta type 9 (large multifunctional peptidase 2) (Psmb9)	-5.31
34	21356	NM_001025313.1	TAP binding protein (Tapbp)	-5.23
49	16913	NM_010724	Proteasome (prosome, macropain) subunit, beta type 8 (large multifunctional peptidase 7) (Psmb8)	-4.38
56	16362	NM_008390.1	Interferon regulatory factor 1 (Irf1)	-4.16
64	21354	NM_013683.1	Transporter 1, ATP-binding cassette, sub-family B (MDR/TAP) (Tap1)	-3.96
65	24088	NM_011905.2	Toll-like receptor 2 (Tlr2)	-3.96

^*Selected expression comparisons between markers from Ikkβ^F/F and Ikkβ^Δhep freshly isolated hepatocytes are listed; the markers with these comparatively most disparate expression levels are clustered within two functional groups.

^**Enrichment, E**, as measured by relative microarray chip hybridization signals, is defined as follows: a negative (−) E value signifies that Ikkβ^Δhep > Ikkβ^F/F, and a positive (+) E value signifies that Ikkβ^F/F > Ikkβ^Δhep (because the initial comparative chip data were expressed as X-fold enrichment of hybridization markers from Ikkβ^F/F compared to those from Ikkβ^Δhep RNA samples).

Western blots

Animals were sacrificed and tissues harvested at 58 and 48 days (see Figures 2A [lanes 3-6] and 2B [lanes 1-4], both panels, respectively), and at 125 days (see Figure 2B [lanes 5 and 6], both panels) of age. Liver extracts (30 μg proteins/lane) and cultured MEF extracts (40 μg) were analyzed on non-reducing prefabricated 10% SDS-PAGE gels (Invitrogen, Carlsbad, CA; Koch et al., 2009). Proteins were transferred to nitrocellulose membranes using an Xcell II blot module (Invitrogen, Carlsbad, CA). Membranes were incubated overnight with either primary (‘1°’) rat anti-mouse CD74 antibody (clone In-1, BD Biosciences Pharmingen, San Diego, CA, catalog (cat.) #555317) or with matched rat IgG2b,κ isotype control (BD Biosciences Pharmingen, San Diego, CA, cat. #553986), washed, incubated 2 hr with secondary (‘2°’) horse radish peroxidase (HRP)-conjugated goat anti-rat IgG (H+L) (Pierce, ThermoScientific, Rockford, IL, cat. #31470), washed, and incubated 1 hr with ECL-Western-Blotting substrate according to the supplier's instructions (Pierce, ThermoScientific, Rockford, IL, cat. #32106). Loading control blots using identical extracts were performed with anti-GRP94 (gp96) antibody (Santa Cruz Biotechnologies, Santa Cruz, CA, cat. #sc-71182). Chemiluminescence was visualized by 5-30 sec exposure to Amersham ECL-Hyperfilm (GE Healthcare Biosciences, Piscataway, NJ, cat. #28-9068-36). Results of Western blots were obtained from different extracts from two independent experiments.

Figure 2.

Detection of CD74 in cell extracts on Western Blots. (A) Ikkβ^F/F and Ikkβ^Δhep mouse livers, and cultured MEFs (each lane from a separate mouse). Lanes: 1 (wildtype MEFs), 2 (Ikkβ^-/- MEFs), 3 (♂Ikkβ^F/F liver), 4 (♂Ikkβ^Δhep liver), 5 (♀Ikkβ^F/F liver), 6 (♀Ikkβ^Δhep liver). Non-specific ∼25 kDa and ∼55 kDa bands in isotype control panels (right upper and lower) likely reflect non-specific cross-reactions with serum or reagent light and heavy IgG chains. (B) Ikkβ^+/+:Alb-Cre mouse livers and spleens, and cultured MEFs (each lane from a separate mouse). Lanes: 1 (♂ #1, Alb-Cre liver), 2 (♂ #1, Alb-Cre spleen), 3 (♂ #2, Alb-Cre liver), 4 (♂ #2, Alb-Cre spleen), 5 (♂Ikkβ^F/F liver), 6 (♂Ikkβ^Δhep liver), 7 (Ikkβ^-/- MEFs), 8 (wildtype MEFs). Loading controls (both panels) showed comparative intensities of 94-96 kDa bands in all lanes.

Immunohistochemistry

Animals were sacrificed and liver tissues (portions of the right lobe) harvested at 58 days of age. Thin sections (5 μm) of paraffin blocks of fixed tissues (4% paraformaldehyde in PBS, 24 hr at 4 °C), and routine hematoxylin and eosin (H&E) stained sections were prepared by standard automated procedures in the UCSD Cancer Center Histology Core (http://cancer.ucsd.edu/Research/Shared/histology.asp). For immunohistochemical staining, endogenous peroxidase activity was blocked in deparaffinized sections with 3% H₂O₂. Sections were pretreated with Retrievagen A* according to the supplier's protocol (BD Pharmingen, San Diego, CA, cat. #550524), followed by 10% fetal bovine serum (Gibco Invitrogen, Carlsbad, CA) to block non-specific binding. Slides were washed, and incubated with 1° anti-CD74 antibody or with matched isotype control rat IgG2b,κ antibody at final concentrations of 5 μg/mL for 1 hr at 21°C. Slides were washed, treated with biotinylated goat 2° anti-rat Ig (BD Pharmingen, San Diego CA, cat. #51-7605KC) for 30 min at 21 °C, washed and treated with Streptravidin-HRP (BD Pharmingen, San Diego CA, cat. #551013) for 30 min at 21 °C. After further washing, the slides were incubated with diaminobenzidine solution (BD Pharmingen, San Diego CA, cat. #551013) for 0.5-5 min (or, as needed), washed, counterstained in Meyer's hematoxylin, dehydrated and coverslipped. Digital images and information (available on request) were acquired with NIS-Elements F 2.30 software, by Nikon DS-Ri1 camera, from Nikon Diaphot with 20× (Fig. 3) or 4× (Fig. 4) objectives. Independent immunohistochemical staining experiments utilizing sections from at least two animals of each genotype were performed multiple times.

Figure 3.

Histochemical studies of CD74 in livers from C57BL/6 ♂, and Ikkβ^F/F and Ikkβ^Δhep ♀ and ♂ mice. Photomicrographs of liver tissue sections stained with H&E, anti-CD74 antibody, or isotype control antibody are shown in the left-, middle-, and right-hand columns, respectively. The sections are from individual C57BL/6 (row 1), Ikkβ^Δhep ♀ (row 2), Ikkβ^F/F ♀ (row 3), Ikkβ^Δhep ♂ (row 4), and Ikkβ^F/F ♂ (row 5) mice. Annotations indicate hepatocyte membrane (small →) and NPC (large →) staining; C, central vein; P, portal triad. (The unannotated H&E stained Ikkβ^Δhep ♀ liver section appears to approach a central region.) Specific CD74 staining is present in hepatocytes from Ikkβ^Δhep mice but not in hepatocytes from C57BL/6 and Ikkβ^F/F mice. Tissue-specific CD74 staining is not seen in sections treated with isotype control antibody. Objective magnification = 20× (the solid bar, lower left, in the C57BL/6 isotype control panel = 200 μm; the scale bar is identical for experimental panels).

Figure 4.

Enriched CD74 staining in Ikkβ^Δhep hepatocytes in midzonal-to-centrilobular regions. Serial sections from a ♂ Ikkβ^Δhep liver, stained with either anti-CD74 or isotype control antibodies as described in the text and in Fig. 3, are here placed in register. Low objective magnification (4×) was used to illustrate the zonal gradient of staining intensity. Hepatocytes proximal to central veins (C) are more intensely stained than hepatocytes proximal to portal triads (P). At lower magnification, scattered NPC staining is more difficult to visualize. The few notations are placed within the same vessels of both serial sections to facilitate registration of the two panels. The solid bar, lower left, in the isotype control panel = 1000 μm (the scale bar is identical for the experimental panel).

Results

Microarray profiling shows upregulation of immune phenotypic expression in Ikkβ^Δhep hepatocytes

Liver tissue and isolated hepatocytes from both genotypes (four separate samples) were compared to see whether profiling differences would be more or less enriched in hepatocytes compared to total liver. RNA sizing and DNA genotyping were performed prior to profiling studies. Intact RNA of high quality (Fig. 1A), and PCR products consistent with wild type and Ikkβ-deleted genomes were observed (Fig. 1B) (Maeda et al., 2003). Microarray hybridizations were performed in triplicate with formatting as diagrammed in Figure 1C.

Key findings in the largest expression group (‘Biological Process’) are shown for freshly isolated livers and freshly isolated hepatocytes in Tables 1 and 2, respectively. Two functional sub-families topped both lists at very high levels of statistical significance (albeit in reverse order): liver (immune system process, p ∼6.5 × 10^-28; defense response, p ∼2 × 10^-27); hepatocytes (defense response, p = 1.4 × 10^-21; immune system process, p ∼1.2 × 10^-18). Examination of the 6-19 genes topping these lists showed that all were expressed at relatively higher levels in Ikkβ^Δhep samples, except for two which were higher in Ikkβ^F/F liver tissue (histocompatibility 2, blastocyst H2-Bl [both sub-families]) and one which was apparently higher in Ikkβ^F/F hepatocytes (CD163 antigen [defense response]).

Enhanced CD74 expression in Ikkβ^Δhep samples ranked highest throughout; CD74 was enriched ∼38-fold in isolated Ikkβ^Δhep hepatocytes, and ∼8-fold in Ikkβ^Δhep liver tissue (a 5-fold enrichment following hepatocyte isolation; see both Tables 1 and 2). Profiling also indicated that additional immune process and defense genes were significantly upregulated >4-10-fold in Ikkβ^Δhep hepatocytes, including MHC class II I-Aα, I-Aβ, and I-Eβ chains, CD44, Cxcl9, guanylate nucleotide binding proteins 1-3, TAP, Tlr2, interferon gamma (IFN-γ) inducible protein 47 and IFN-γ regulatory factor 1 (see both Tables 1 and 2). Direct inspection of the microarray data also revealed augmented expression levels of Ikkβ^Δhep hepatocyte CIITA, STAT1, cathepsin S, co-activator CD86 and programmed death ligand-1 (PD-L1) (> 6-, 5.4-, 3.6-, 2- and 5.5-fold, respectively), all significantly greater than isolated liver tissue. As shown in Table 3, the available data from microarray comparisons between Ikkβ^Δhep hepatocytes and Ikkβ^-/- MEFs revealed few similarities (Chen et al., 2006).

Table 3.

Microarray profiling: Partial comparisons between mouse embryo fibroblasts (MEFs, Ikkβ^-/- vs. wild type MEFs) and freshly isolated hepatocytes (Ikkβ^Δhep vs. Ikkβ^F/F).

Gene	MEFs (Ikkβ-/- vs. wt)	Hepatocytes (IkkβΔhep vs. IkkβF/F)	GenBank Acc. #	Function
	X-fold enrichment*
CARP	+156.3	-1.21	AF041847	Cardiac ankyrin repeat protein, binding to myopalladin
Versican	+9.7	NR	D45889	Anti-cell adhesive molecules
α-Actin	+2.6	+1.28	X13297	Cytoskeleton, stress fibers
Prothymosin β4	+2.34	-1.98	U38967	Prevents actin polymerization
S100A4	+2.2	NR	M36579	Tumor metastasis, calcium-binding protein
Tropomyosin 2	+2.0	-1.47	M22479	Actin-binding protein
IGFBP2	-134	-1.79	X81580	Inhibitor of cell mitogenesis
TIMP-2	-5.8	-2.32	X62622	Inhibitor of matrix metalloproteinases
TIMP-3	-4.0	-1.99	U26437	Inhibitor of matrix metalloproteinases
VCAM	-2.9	+5.62	M84487	Vascular cell adhesion molecule
Integrin β5	-2.0	-1.2	AF022110	Cell adhesion

X-fold enrichment values for MEFs were taken from Chen et al., 2006. The X-fold enrichment value is here in Table 3 defined as the relative comparison of RNA hybridization expression of MEF Ikkβ^-/- to MEF Ikkβ^+/+, or of Ikkβ^Δhep to Ikkβ^F/F such that a positive (+) value signifies that MEF Ikkβ^-/- > MEF Ikkβ^+/+, or that Ikkβ^Δhep > Ikkβ^F/F; and, a negative (−) value signifies that MEF Ikkβ^+/+ > MEF Ikkβ^-/-, or that Ikkβ^F/F > Ikkβ^Δhep. Note that this definition is here the reverse of that in Tables 1 and 2, in order to establish congruence with the E values previously published for the MEF Ikkβ^-/- and Ikkβ^+/+ pair.

Ectopic expression of CD74 proteins is upregulated in Ikkβ^Δhep hepatocyte extracts

Four specific bands (M_r ∼ 31, 35, 41 kDa, and a 10-12 kDa doublet) were detected in Ikkβ^Δhep liver extracts by antibody specific to cytosolic CD74 epitopes (Figs. 2A and 2B [left panels]). CD74-specific bands were not seen in similar extracts probed with isotype control antibody (Figs. 2A and 2B [right panels]). Band intensities were similar in ♂ and ♀ samples; the relative intensities of p31 were ∼2.5-fold higher than p41 as expected (Bertolino and Rabourdin-Combe, 1996). In contrast, some (p31 and p41) but not all (p35, p45 and the 10-12 kDa doublet) CD74-specific bands were detectable at significantly lower levels in Ikkβ^F/F liver extracts (Figs. 2A and 2B [left panels]); these bands were likely due to CD74 expressed by non-parenchymal cells (NPCs) in the intact livers (Fig. 3). The finding of CD74-specific expression in Ikkβ^Δhep liver extracts was not an artifact of albumin promoter-driven Cre expression, as CD74 expression levels in liver extracts from Ikkβ^+/+:Alb-Cre mice were similar to those from Ikkβ^F/F mice, with no detectable p35, p45, or 10-12 kDa doublet (Fig. 2B).

As expected from profiling (Chen et al., 2006), CD74 was undetectable in wildtype MEF extracts whereas extracts from Ikkβ^-/- MEFs variably expressed very low levels of p31 and p41 bands but undetectable levels of p35 or the 10-12 kDa doublet (Figs. 2A and 2B [left panels]). The significance of these observations is currently unclear.

Ectopic and specific expression of CD74 is observed in situ in Ikkβ^Δhep hepatocytes

To localize hepatic CD74 expression, livers from ♂ C57BL/6, and ♂ and ♀ Ikkβ^F/F and Ikkβ^Δhep mice were analyzed by routine histology and immunohistochemistry. In all cases, the same 1° antibody used for Western blots was employed; with respect to Ikkβ^F/F and Ikkβ^Δhep mice, portions of the same tissue samples used for Western blots (Fig. 2A) were fixed for histochemistry (Figs. 3 and 4).

Normal histology was observed by H&E staining of C57BL/6 liver (Fig. 3, row 1). As expected (Maubach et al., 2007; Momburg et al, 1986), specific CD74 staining was observed in NPCs bordering a portal triad, and also scattered sparsely throughout the liver. In contrast, specific staining was not observed either in hepatocytes, or in cells bordering the central vein, or in sections incubated with isotype control antibody (Fig. 3, row 1). The faint CD74-positive p31 and p41 bands on Western blots of liver extracts from Ikkβ^+/+:Alb-Cre and Ikkβ^F/F mice (Figs. 2A and 2B) may reflect NPC staining. Further work is needed for precise identification of NPCs, whether dendritic, HSC, Kupffer, sinusoidal endothelial and/or biliary ductular epithelial cells, which are responsible for this CD74-positive staining.

As shown in Fig. 3 (rows 2-5), H&E staining showed no morphological differences between Ikkβ^Δhep, and control Ikkβ^F/F, ♀ or ♂ livers. However, intense CD74 staining was observed in virtually all Ikkβ^Δhep hepatocytes throughout liver acini and regardless of gender. Specific CD74 staining in Ikkβ^Δhep hepatocytes was associated with cytoplasmic and with membrane sites. In contrast, although scattered periportal and acinar staining was observed in NPCs, specific CD74 staining was undetectable in Ikkβ^F/F hepatocytes. No CD74 staining was observed in any liver sections incubated with isotype control antibody (Fig. 3) or in the absence of 1° antibody (data not shown).

By observation of liver sections at lower magnification (Fig. 4), specific CD74 staining in Ikkβ^Δhep livers appeared to be more intense in hepatocytes bordering central veins, and to decrease slightly and gradually in intensity from centrilobular to periportal zones.

Discussion

Immunohistochemical, microarray profiling, and Western blot findings show ectopic and gender-independent expression of major, minor and putatively processed CD74 isoforms in hepatocytes deleted in Ikkβ (Ikkβ^Δhep). Immunohistochemistry suggests that antibody-specific CD74 expression occurs in all hepatocytes, as might be expected from Ikkβ-deletion analyses indicating no undeleted product in a population of isolated Ikkβ^Δhep hepatocytes (Maeda et al., 2003; Koch et al, 2009); however, the isoform(s) responsible for the observed immunochemical reactivities have not been specifically identified. The prominent 35 kDa band in Ikkβ^Δhep liver extracts might represent another CD74 murine isoform, so far unreported; the nature of the 45 kDa band observed in younger but not older mice (compare Fig. 2A with Fig. 2B) is unclear. The 10-12 kDa doublet, also observed in spleens from Ikkβ^+/+:Alb-Cre mice (Fig. 2B [left panel]), might be a form of RIP-processed CD74 (Becker-Herman et al., 2005); if functional, this might account for residual NF-κB activity in Ikkβ^Δhep hepatocytes (Maeda et al., 2003).

The occurrence of multiple isoforms complicates interpretation of immunohistochemical observation of enriched midzonal-to-centrilobular localization of hepatocellular CD74, since site-enrichment could be due either to increased expression of all or selected isoforms, or to increased availability of antigenic sites of one or more equally abundant isoforms. Alternatively, owing to the relatively short paraformaldehyde fixation time, and to subsequent variably effective antigen retrieval steps, zonal staining intensity differences might reflect differential fixation of antigenic sites accessible to the anti-CD74 antibody employed. Whereas longer fixation times (> 24 hr) clearly reduce antigen detection (data not shown), zonal differences were still visible. Therefore, because the zonal differentiation of hepatocytes is responsible for many functional attributes of liver under both normal, and pathological conditions, further experiments are needed to clarify these observations as well as to eliminate other possible fixation artifacts.

Although the upregulation of CD74 expression in hepatocytes appears to be a consequence of targeted hepatocellular deletion of Ikkβ in Ikkβ^Δhep mice, Ikkβ deletion alone is not a uniformly sufficient condition of CD74 upregulation, because Ikkβ^-/- MEFs express only barely detectable levels of CD74, as revealed by Western blots (Fig. 2), and none by RNA profiling (Chen et al., 2006). While the mechanisms responsible for ectopic hepatocellular CD74 expression remain to be determined, they do not appear to be random or isolated. This conclusion is supported by liver micorarray profiling results which reveal, with very high statistical significance, that augmented CD74 expression in Ikkβ^Δhep hepatocytes is concomitant with enriched expression of two large families of immune system-related genes, among them CD44, MHC class II I-Aα, I-Aβ, and I-Eβ chains, CIITA and PD-L1 mRNAs, the functions of which are involved in antigen processing, host defense and liver tolerance (Tiegs and Lohse, 2009). This enrichment is selectively greater in isolated hepatocytes compared to that in unfractionated liver extracts. Although further work is needed to show the translation of these mRNAs directly and to characterize the cellular locations of these putatively translated proteins, these additional findings are surprising and intriguing because the expression of all five of these molecules, like CD74, is usually restricted to lymphoid (Abbas et al., 2010; Rafi-Janajreh et al., 1998) and NPCs (Tiegs and Lohse, 2009).

The collective, and significantly augmented, expression of CD74 and these two gene families in Ikkβ^Δhep hepatocytes suggests several interrelated hypotheses about ectopic immunologic functions and susceptibility to HCC in these mice, as discussed elsewhere (Koch and Leffert, 2010).

Abbreviations

Alb

refers to albumin-specific gene promotor

BLTLF

Bacille-Calmette-Guerin-primed lipopolysaccharide-induced T cell-mediated acute liver failure

ConA

lectin concanavalin A

CIITA

transcriptional coactivator of MHC class II expression

Cre

cre recombinase

DEN

diethylnitrosamine

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

H&E

hematoxylin and eosin

HRP

horse radish peroxidase

HSC

hepatic stellate cell

IFN-γ

interferon γ

Ii

invariant chain CD74 chaperone

IKK

IκB complex

IKKβ

protein kinase subunit of IKK

Ikkβ^F/F

mouse strain carrying Cre-recombinase sites (F/F) spanning Ikkβ exon 3

Ikkβ^Δhep

mouse strain Ikkβ^F/F:Alb-Cre, specifically deleted in hepatocyte Ikkβ

IL

interleukin

JNK

c-JUN nuclear kinase

LPS

lipopolysaccharide

MAPK

mitogen-activated protein kinase

MEFs

mouse embryonic fibroblasts

MHC

major histocompatibility

MIF

macrophage migration inhibitory factor

NPC

non-parenchymal cell

NF-κB

nuclear factor kappa B

PD-L1

programmed death ligand

RIP

regulated intramembrane proteolysis

ROS

reactive oxygen species

STAT

signal transducers and activator of transcription

TNF

tumor necrosis factor

TNFR2

tumor necrosis factor receptor 2