W ith the development of | monoclonal antibodies (mAh) to individual HLA determinants it become possible to distinguish between HLA-DP, -DQ and -DR and to study MHC class II distribution in detail. The expression of MHC class II on lymphoid cells is a dynamic process. MHC class II expression can vary with physiological compartment, stage in differentiation and state of activation - all determining factors influencing T cell development and response. In most cases, however, MHC class II expression by many cell types (including B cells, macrophages and dendritic cells) is low or absent unless induced by some stimulus, usually cytokines (Houghton et al.,
1984; Collins et al., 1984; Pober et al., 1983).
Modulators of MHC class II expression
Many cytokines are capable of inducing MHC class II expression (Balkwill and Burke 1989). They generally act on specific cells and can function independently or in concert with other cytokines. The molecular response to fFN-yis the best studied. The latter is a potent inducer of MHC class II expression in macrophages and cells of the non-haemopoietic lineage (Houghton et al., 1984; Collins et a l, 1984; Pober et a l,
1983). It is generally accepted that IFN-y causes a specific rise in MHC class II mRNA levels, which begins ~4-8 hours after addition and reaches a maximum by 24 hours (Collins et al., 1984). A lag in cell surface expression means that maximal levels are detected after 2 days. The genetic elements involved in this induction will be outlined later in section 1.3. For B cells IL-4 is the most potent induction stimulus (Roehm et al., 1984; Rousset et al., 1988), increasing the cell surface levels of MHC class II within 48-72 hours (Noelle etal., 1986). IL-4 can increase the level of class II on resting B cells as much as 10 to 15 fold, and this induction is relatively specific as the levels of Ig and MHC class I are only slightly increased (Noelle et al., 1984). The IL-4 mediated induction of MHC class II can be down-regulated by IFN-y (Mond et
al., 1986). Other inducers include TNF-a and GM-CSF for monocytes and IL-10 for B cells (Pfizenmaier et at., 1987; Alvaro-Garcia et at., 1989; Go et al., 1990) In general TN F-a acts in synergy with IFN-y at the level of transcription (Arenzana- Seisdedos eta l., 1988). In addition, recent evidence suggests that while TN F-a has a stimulatory effect early after treatment it can also be inhibitory with prolonged treatment (Zimmer and Jones 1990). A number of other agents have also been shown to downregulate class II expression. IFN-y is an antagonist for IL-4 in B cells (Rousset et al., 1988), whilst glucocorticoids are inhibitory for expression in both macrophages and B cells (Dennis and Mond 1986). Prostaglandins may also be inhibitory under certain circumstances (Snyder etal., 1982).
Evidence for selective expression of class II MHC isotypes
There have been few attempts to analyse the expression of the individual MHC class II isotypes on specific cell populations, most work focusing on the expression of HLA- DR as being representative of all human class II molecules. However, MHC class II genes are never expressed equally. The different levels of expression of HLA-DR, -DQ and -DP occur during haemopoietic progenitor cell differentiation. HLA-DQ expression, in particular, is not always concordant with the other HLA class II molecules expressed early in progenitor cells. The regulatory mechanisms which control the locus specific expression of the class II genes are not fully understood, however, such discordant expression of DQ implies that regulatory mechanisms which are specific to this isotype must exist. The following is an outline of isotype specific expression of MHC class II molecules on various cell types.
The tissue distribution of HLA antigens characteristic of the adult is gradually attained during ontogeny. The maturation process in man is complete after 32 weeks of gestation (Natali et al., 1982), the tissue distribution reached during foetal life being maintained in adults (unlike MHC class I). The tissue distribution and ontogeny of
HLA-DQ differs from that of -DR: the former is confined to thymic epithelial cells and to endothelial cells of the small intestine during the first 26 weeks of intrauterine life, whilst -DR attains its full tissue distribution after 26 weeks (Natali et al,, 1984). Studies in mice have revealed that the major difference between the two species is the appearance of MHC class 11+ epidermal dendritic cells; in man these appear in intrauterine life, whilst in mice they are only detectable in 5-day old neonates (Natali et al,, 1984). In both species, as well as in rabbits, rats and guinea pigs MHC class 11+ cells are first detected in the thymus at an early stage (Rouse et al,, 1979; Jenkinson et al,, 1980; Jenkinson etal,, 1981; Natali etal,, 1981a,b).
The expression of MHC class II molecules on cells of the B cell lineage is regulated throughout their developmental pathway, as previously mentioned. Immature pre-B cells usually lack HLA-DQ expression (-20% +) whilst being positive for -DR (Fermand etal., 1985). In some studies the acquisition of surface IgM was associated with the concurrent acquistion of HLA class II, with the density of the two molecules increasing in parallel; in other studies the two molecules appeared to be regulated independently (Lala et ah, 1979; Dasch and Jones, 1986). As the cells mature HLA- DQ expression increases, so that over 95% of B cells from blood and lymph carry -DQ and -DR. Mature B cells express HLA class II constitutively, one study estimating the levels to be 3x10^ "DR, 3x10^ -DQ and 2x10"^ -DP molecules per cell (Mond etal,, 1980; Greenstein etal,, 1981). However considerable heterogeneity can be observed in cells at different stages of development and may reflect functional differences amongst these sub-sets. Stimulation leads to an alteration in the cell phenotype, over 70% of activated B cells losing HLA-DQ expression, whilst retaining low levels of -DR expression. Terminal differentiation into plasma cells is accompanied by a total loss of MHC class II expression (Latron et al,, 1988; Dellabona et al,, 1989).
In the myelomonocytic lineage, MHC class II molecules have been detected on monocytes (Winchester eta l.,191^), normal myeloblasts (Winchester etal., 1977) as well as committed stem cells for granulocytes and macrophages (Cline and Billing, 1977; Sieff etal., 1982). The earliest cells showing monocyte/macrophage markers are detected at week 4-6 of fertilisation and at this stage are HLA-DR"; as the cells seed to the thymic cortex, lymph, spleen and bone marrow they become -DR+. Interestingly, HLA-DQ molecules are not detectable on the stem cells of any lineage, including those destined to differentiate into monocytes. As with B cells there appears to be a progression from HLA-DR+/DQ" to -DR+/DQ+ as the macrophages mature, terminally differentiated Langerhans and Kupffer cells for example expressing both HLA-DR and -DQ antigens (Janossy et al., 1986).
MHC class II expression by T cells shows considerable inter-species variation, unlike B cells and dendritic cells which constituitively express MHC class II molecules in all species studied so far (Kaufman et al., 1984; Kappes and Strominger 1988). Resting T cells are generally MHC class II negative and reports of extremely low numbers of HLA-DR+/DP+/DQ" resting T cells may be due to the presence of a few activated cells (Brown et al., 1984). Natural killer cells do not express detectable levels of MHC class II molecules. In humans resting T cells are negative whereas cells activated with anti-CD3 mAh have been shown in one study to express -1x10"^ HLA- DR, SxlO'^'DQ and 5xlO^ DP molecules per cell, however, the levels are still significantly below that of other human accessory cell populations (Robbins et al.,
1988). In a study by Oshima and Eckels, (1990) using two colour FACS analysis the expression of HLA-DR, -DQ and -DP class II molecules on T cells was analysed after activation by allogeneic B-LCL. A dramatic and sustained increase in all isotypes was observed over 6-7 days, however, a hierarchy of expression was apparent with HLA- DR expression always being the highest, -DQ being the lowest and -DP intermediate. At day 8 three populations were observed, one which was HLA-DR+, -DP+ and -DQ+, and two populations which were completely -DQ negative. Interestingly no
H LA -D R" populations were observed. These patterns of HLA class II isotype expression were similar in CD2+, CD4+ and CD8+ subgroups of T cells. All equine and rat T cells express MHC class II molecules upon activation whereas in mice the evidence is controversial (Lorber et al., 1982; Singh et a l, 1984; Crepaldi et at.,
1986). Early work using cloned lines of alloreactive cytolytic and helper T cells which were examined for the expression of H-2A molecules using FACS analysis suggested that alloreactive mouse T cells could adsorb MHC class II molecules from the stimulating cell population (Lorber et at., 1982). However, evidence does exist that points towards the presence of endogenously synthesised MHC class II molecules on activated T cells (Evans et al., 1978; Ko et al., 1979). HLA-DR, -DQ and -DP molecules are differentially expressed on T cells after stimulation (Robbins e ta l.,
1988). In addition, the level of expression would appear to vary with stage of cell cycle. In recent years the most comprehensive study of MHC class II expression on T cell sub-sets during maturation has been that of Dutia et al., (1993) and Hopkins et al.,
(1993), although this focuses on sheep it gives an indication of the sequence of events. The results of this study showed that the expression of MHC class II by T cells is dependant upon the age and location of the cell. In foetal blood CD8+/yô cells were HLA class II negative and the expression on CD4+ cells was very low. As the cells mature they progress from being HLA-DQ"/DR" to -DQ"/DR+ and then -DQ+^DR^ so that all blast cells expressed a uniformly high level of -DQ and -DR with no differential expression of the class II products. Within the afferent lymph and node HLA-DQ+/DR+ cells typified recently activated cells; whereas resting and recirculating memory T cells in the efferent lymph and node were -DQ"/DR+
MHC class II expression in the thymus
Little is known of the differential expression of the various MHC class II isotypes in the thymus. Early work by Natali et ah, (1984) and Takenouchi and co-workers (1986) indicated that HLA-DQ antigenic expression was more limited than that of
-DR. Ishikura et al., (1987) attempted to clarify the thymic expression of these antigens by using well characterised mAbs They demonstrated that there were different distributions of these antigens in the human thymus. In the cortex both HLA-DR and -DQ were strongly expressed on thymic epithelial cells, with simultaneous expression of both antigens. On the other hand in the medulla HLA-DQ antigens had a more limited pattern of expression than -DR. Double staining revealed small numbers of dual positive (HLA-DR+"DQ+) cells in the medulla, which looked like thymic epithelial cells (TEC), indicating that at least some medullary TEC were double positive, although the presence of small numbers of HLA-DR'DQ“ and -DR+DQ" cells could not be excluded. The preponderance of HLA-DR over -DQ was consistently found in ex vivo tissues from individuals aged between 10 weeks - 2 years and could be attributed to very weak/no expression of -DQ antigens on dendritic cells and/or macrophages.
Furthermore, evidence now exists for isotype-specific distribution of an MHC class II molecule (H-20) in the mouse thymus. Karlsson etal., (1991) identified a novel class II molecule when they precipitated H -20 (previously H-2Ap2). Electrophoretic analysis demonstrated that the H-20P chain pairs with a novel a chain and that the heterodimer associates with the invariant chain. Preliminary work indicated that H -20 had a more limited thymic distribution than H-2A. H-2A was present throughout the medulla and on cortical epithelial cells. H-20p, in contrast, was limited to aggregates of cells in the medulla. Assessing the distribution of MHC molecules on thymic epithelium is difficult due to ambiguous results caused by staining of the dense network of MHC class 11+ bone-marrow derived cells, however, this problem can be overcome by examining a bone marrow chimeric thymus. In this case all the bone- marrow derived cells are of donor origin and the host MHC class II is restricted to the epithelial cells. Using bone marrow chimeras and MHC transgenic mice Surh et al
(1992) demonstrated that H -20 was expressed on B cells but, unlike 'normal' MHC class II, was not detectable on macrophages or dendritic cells. In addition, not only was
H -20 expression confined to the medulla but, cells which were positive for H -20 had little or no expression of H-2A or MHC class I molecules. Therefore, the expression of a 'typical' MHC class II molecule (H-2A) and H -20 would appear to subdivide the medullary epithelium into two phenotypically distinct subsets in adults.
I Capsu^
H assa I â _
corcuscie
MedJiia/y
epreliurr
Figure 1.5: Cellular composition of the thymus
Schematic representation of the stromal cell populations within the thymic medulla and cortex. All the intervening spaces are filled with lymphocytes (not shown).
Table 1.0: Phenotypic classification of the thymic micro-environment
Cell type Phenotype R eferences
Cortical epithelial cells MHC class II hi Boyd et a l, 1977; de Maagd et a l, 1985; Wekerle et a l, 1980
Medullary epithelial cells MHC class n variable Bofill e ta l, 1985; F a.netal, 1983; Jenkinson et a i, 1981; Surh et
a l, 1993; van Ewijk et a l, 1980; von Gaudecker et a l, 1986
Macrophages MHC class n hi Boyd e ta l, 1993
Dendritic cells/ Interdigitating cells
MHC class n hi Boyd et a l, 1993
B cells MHC class II + Boyd e ta l, 1993