Spleen-specific Expression of the Malaria-inducible Intronless Mouse Gene imap38 *

We characterize the mouse gene imap38and its inducibility by Plasmodium chabaudi malaria among different lymphoid tissues and mouse strains of differentH-2 complex and non-H-2 background.Imap38 is a single copy gene assigned to chromosome 6B. It consists of only one exon of 1900 base pairs encoding a highly basic 25.8-kDa protein. Confocal laser scanning microscopy localizes differently tagged IMAP38 proteins in nuclei of transfected cells. Reporter gene assays reveal that the 1730-base pair 5′-flanking region, containing an RSINE1 repeat immediately adjacent to initiation site +1, exhibits promoter activity in nonmurine cells, while it is largely repressed in diverse mouse cell lines, which corresponds to the situation in mouse tissues. P. chabaudi malaria inducesimap38 expression almost exclusively in the spleen but not in other lymphoid organs. Parasite lysates are able to induceimap38 in the spleen, but not in spleen cells ex vivo. Activation of spleen cells by LPS and other stimuli is not sufficient to induce imap38. Inducibility ofimap38 requires signals from both parasites and the intact spleen, and it is controlled by genes of that non-H-2background, which also controls development of protective immunity against P. chabaudi malaria.

Natural immunity to malaria is known to be predominantly directed against the blood stages of the causative agent, parasite protozoans of the genus Plasmodium. Remarkably, however, this immunity is unable to prevent parasitemia during malaria season, but it can completely suppress disease symptoms (1,2). The spleen plays a central role in immunity against blood stage malaria (3)(4)(5). It is the major site of (i) elimination of parasite-infected erythrocytes via erythrophagocytosis, (ii) elaboration of protective immune mechanisms, and (iii) hypersensitivity reactions manifesting themselves as spleen enlargement (6), not infrequently resulting in fatal spleen rupture (7). However, the events occurring in the spleen in context with survival and disease are not yet really understood.
A convenient model to study the role of the spleen in malaria is the murine malaria Plasmodium chabaudi, which shares several common characteristics with the most dangerous hu-man parasite P. falciparum, causing malaria tropica (8). Blood stage infections with P. chabaudi take a self-healing course in female C57BL/10 mice, which subsequently become immune against homologous rechallenge (9). Self-healing and, thus, the development of protective immunity is controlled by genes of the H-2 complex, genes of the non-H-2 background, and gender and testosterone, respectively (10,11). The spleen dramatically increases in size during acute malaria, and it remains enlarged in immune mice. We have shown that spleen enlargement and development of protective immunity in C57BL/10 mice correlate with expression of the novel gene iap38 (12), recently renamed imap38. Noninfected mice express only very low levels of imap38, whereas its expression is increased by about 50-fold in the spleen of P. chabaudi immune mice. The imap38 gene remains constitutively expressed at such a high level in the spleens of immune mice even after 13 weeks, when parasites have already been eliminated for more than 8 -10 weeks. Only the cDNA of imap38 has been preliminary characterized to date. This study is aimed at characterizing the imap38 gene and its inducibility in more detail.

MATERIALS AND METHODS
Mice and Infections-Mice of the inbred strains C57BL/10, B10.A, B10.D2, C57BL/6, BALB/c, and DBA2/J were bred under specific pathogen-free conditions in our animal facilities. P. chabaudi malaria was maintained in NMRI mice by weekly passage as described previously (11). P. chabaudi-parasitized erythrocytes (1 ϫ 10 6 /500 l of PBS 1 ) were intraperitoneally injected in mice. Parasitemia and cell number were determined as described previously (11). Immune mice were those at 9 weeks after infection with P. chabaudi (13).
Southern Blot Analysis-Genomic DNA isolated from kidneys of female C57BL/10 mice was digested with restriction enzymes, separated in 0.7% agarose gels, and blotted by a downward alkaline transfer method (14). Hybridizations were carried out using the Ecl136II/EcoRIfragment (positions 2185-3240) of the cDNA as a probe as described recently (12).
Northern Blotting-RNA was isolated from spleen, other lymphoid tissues, and cells (15). Northern blots using 20 g of glyoxylated total RNA were performed as described recently (12). Evaluation of autoradiograms was done by densitometric scanning with Quantiscan software (Biosoft, Cambridge, UK).
Panhandle Vectorette PCR-Libraries for amplification of flanking sequences were constructed as described by Siebert et al. (16). After digesting genomic DNA with AluI, Bsh1236I, Ecl136II, PvuII, RsaI, or SmaI, adapters were ligated to the blunt ends. The first PCR contained 0.3 M Ap-1 (5Ј-GGATCCTAATACGACTCACTATAGGGC-3Ј) and Walk-1a primer (5Ј-CTCTCCTTTCTGCATTTGGATTCTCTGTGC-3Ј), 0.2 mM dNTPs, and 3.75 units of Expand High Fidelity DNA-polymerase mix (Roche Molecular Biochemicals) in 50 l of 1ϫ Expand High * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AJ133125.
Fidelity buffer with Mg 2ϩ . A second round of PCR was performed with the nested primers Ap-2 (5Ј-CTCACTATAGGGCTCGAGCGGC-3Ј) and Walk-1b (5Ј-GCAGATGCTCCTGCTGACGTTTCACAAAAT-3Ј) and 1 l of a 1:100 dilution of the first reaction as template. Both PCR reactions were programmed 2 min at 94°C, 35 cycles of 15 s at 94°C, 8 min at 68°C, and a final extension step at 72°C for 10 min. Products were analyzed on agarose gels; bands were eluted and cloned into pMOSBlue (Amersham Pharmacia Biotech, Freiburg, Germany).
Inverse PCR-Genomic DNA digested with TaqI was circularized at a final concentration of 3 g/ml (17) with T4 DNA ligase (Fermentas, St. Leon Roth, Germany), and 3-l aliquots were used as template for PCR under conditions as described before with primers Walk-1a and Walk-2a (5Ј-GTATGAGCTGGTGCAGGACACGCGGTGCGCTGACC-3Ј). Then 1 l of the first reaction was reamplified with Walk-1b and Walk-2b (5Ј-TACTGGAAGGGCTGGAGGCGTGGTTTCTCTGTCTT-3Ј). The PCR product was eluted from agarose gels and directly sequenced.
Screening of a Cosmid Library-The 129/ola mouse cosmid library (library no. 121) constructed from genomic DNA of 129/ola mice was screened using the Ecl136II/EcoRI fragment as probe by the service facilities of the German Human Genome Project (Heidelberg). Restriction fragments of a positive clone were subcloned into pBluescript (Stratagene, Heidelberg, Germany).
DNA Sequencing and Computer Analysis-Sequence reactions were performed with the Fluorescent-Labeled-Cycle-Sequencing kit (Amersham Pharmacia Biotech) using infrared fluorescence-labeled primers (MWG Biotech, Ebersberg, Germany) and analyzed in a LICOR4000 DNA sequencer. The programs FASTA (18), BLITZ (19), and BLAST (20) were used to search EMBL and Swiss-Prot data bases. Protein sequences were analyzed with SAPS (21), ScanPROSITE, and Prot-Param software (22). Repetitive elements in DNA sequences were identified using the program CENSOR (23). The TRANSFAC data base (24) was searched using the program MatInspector (Genomatrix, Mü nchen, Germany).
Fluorescence in Situ Hybridization-Chromosomal mapping was performed according to a previously described protocol (25). 80 ng of the biotin-labeled cosmid DNA (MPGc121K124460Q2) were combined with 3 g of mouse Cot1 DNA and 7 g of salmon sperm DNA in a 12-l hybridization mixture. Mouse metaphase chromosomes were prepared from spleen cells of a female Balb/c mouse following a standard procedure (26). Following hybridization for 17 h at 37°C and posthybridization washes, the hybridized probes were detected via FITC-conjugated avidin. Chromosomes were banded by staining with 4,6Ј-diamino-2phenylindole dihydrochloride. Digitized images of emitted 4,6Ј-diamino-2-phenylindole dihydrochloride and FITC fluorescence were recorded separately using a CCD camera (Photometrics), carefully aligned, and electronically overlaid.
Mapping of Transcription Starts-Reverse ligation-mediated rapid amplification of cDNA ends was used according to a modification of protocols previously described (27)(28)(29). Poly(A ϩ )-RNA (10 g) from spleens of immune female C57BL/10 mice was dephosphorylated with 3 units of shrimp alkaline phosphatase (Amersham Pharmacia Biotech) at 37°C for 2 h. After heat inactivation of the enzyme and precipitation of the RNA, the pyrophosphate bond in the 5Ј-cap was hydrolyzed by incubation with 10 units of tobacco acid pyrophosphatase (Epicentre, Hess. Oldendorf, Germany) in 10 l at 37°C for 2 h before reprecipitation. The RNA adapter 5Ј-CUAAUACGACUCACUAUAGGGCUCGAG-CGGCCGCCCGGGCAGGU-3Ј was ligated to the 5Ј-ends at 16°C for Reactions were incubated at 22°C for 15 min, at 55°C for 60 min, and at 95°C for 5 min. The following PCRs used the primers described above for panhandle vectorette PCR. 25 l of 1ϫ RT buffer with Mg 2ϩ containing 0.4 M Ap-1 and Walk-1a, 0.2 mM dNTPs, and 3.75 units of Expand High Fidelity enzyme mix were added, and, after a 2-min denaturation at 94°C, 35 cycles were performed for 15 s at 94°C, 4 min 68°C, and a final extension of 10 min at 72°C. A nested PCR was performed with 1 l of the first reaction as template and primers Ap-2 and Walk-1b. PCR products were purified with a PCR purification kit (Qiagen, Hilden, Germany) and cloned into pMOSBlue, plated on isopropyl-1-thio-␤-D-galactopyranoside and 5-bromo-4-chloro-3-indolyl-␤-D-galactopyranoside-containing LB/Amp/Tet agar plates, and 40 white clones were picked and sequenced.
In Vitro Transcription and Translation-The TNT T7/T3 coupled reticulocyte lysate system (Promega) was used to transcribe and translate the plasmids iapG2 and iap⌬289 in the presence of 35 S-labeled cysteine and methionine (ICN, Eschwege, Germany) according to the manufacturer's protocol. The reaction products were separated by SDSpolyacrylamide gel electrophoresis (30) and fluorographed using Amplify fluorography solution (Amersham Pharmacia Biotech) and Kodak BioMax MR film with intensifier screen.
Confocal Laser Scanning Microscopy-The vector pMyc6ϫHis was derived from pSecTag A (Invitrogen, Leeg, NL) by digestion with SfiI and NheI, followed by blunting and religation to eliminate the secretion signal. The coding region and the 5Ј-UTR of imap38 were amplified from clone iapG2 using the upstream primer IAPGes-1 and one of the three downstream primers IAP-frameA (5Ј-TGACTCGAGGCCTGAGC-CTCGCGCTCGCTGCCTGTG-3Ј), IAP-frameB (5Ј-TGACTCGAGAGC-CTGAGCCTCGCGCTCGCTGCCTGTG-3Ј) or IAP-frameC (5Ј-TGACT-CGAGAAGCCTGAGCCTCGCGCTCGCTGCCTGTG-3Ј). PCR products were digested with HindIII and XhoI, cutting in the 5Ј-UTR and the downstream primer, respectively, and inserted between the HindIII and XhoI sites in pMyc6ϫHis. The plasmids were designated pIMAP-MycA, -B, and -C. The plasmid pEGFP-IMAP was constructed by amplifying the coding region of imap38 with the primers 5Ј-GGAAGATC-TATGCAGAAAGGAGAGACG-3Ј and 5Ј-CTAACACGCTAGCGGTCTG-TTATTCTGCTCCCC-3Ј. The PCR product was digested with BglII and EcoRI, cutting in the upstream primer and the 3Ј-UTR, respectively, and inserted between the BglII and EcoRI sites of pEGFP-C1 (CLON-TECH, Heidelberg), generating a continuous EGFP-IMAP38 ORF. CHO-K1 cells grown on coverslips in 35-mm dishes were transfected with 1 g of plasmid and 3 l of FUGENE transfection reagent (Roche Molecular Biochemicals). Cells were cultured for 24 h, washed twice with PBS, and fixed with 3.7% formaldehyde in PBS. After thorough washings, detection of Myc-tagged proteins by immunofluorescence and Western blotting was performed according to a recently published protocol (31). For immunofluorescence, a 1:100 dilution of anti-c-Myc (A14) (Santa Cruz Biotechnology, Heidelberg) and anti-rabbit IgG FITC conjugate (working dilution 1:80; Sigma) were used as primary and secondary antibodies. Coverslips were incubated with RNase A (10 g/ml in PBS) at 37°C for 1 h and then mounted on slides in a 1:1 (v/v) mixture of glycerol and Vectashield (Serva, Heidelberg) containing 2% (w/v) 1,4-diazobicyclo-(2,2,2)octane (Merck, Darmstadt, Germany) and 5 g/ml propidium iodide. FITC and propidium iodide fluorescence were analyzed using the confocal laser scanning microscope Leica TCS NT version 1.5.451 (Leica Lasertechnik, Heidelberg) after excitation with the 488-nm argon laser line. Optical sections of 0.5-m intervals were evaluated using Adobe Photoshop 5.0 and Corel Draw 8. For Western blotting, the anti-c-Myc antibody (A14) was used at a working concentration of 0.2 g/ml. Horseradish peroxidase-conjugated goat anti-rabbit IgG (Santa Cruz Biotechnology) diluted 1:25,000 in TST (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.05% Tween-20) was used as a secondary antibody.
Preparation of P. chabaudi Lysates-Blood was obtained from mice with parasitemia higher than 40%. Cells were washed twice with PBS, and parasites were lysed by three freeze-thaw cycles between Ϫ80 and 37°C followed by treatment with an ultrasound microtip (Sonifier B12, Branson Sonic Power Company, Danbury, CT) three times for 5 s. Contol lysates had equivalent concentrations of total erythrocytes but were prepared from uninfected mice. Extracts were stored at Ϫ20°C.
Spleen Cell Conditioned Medium-Spleens were aseptically removed from B10.A mice, and single cell suspensions were prepared as described previously (13). Cells were seeded at a density of 5 ϫ 10 6 /ml in culture medium containing 2 g/ml concanavalin A (Sigma) and cultured for 48 h. Supernatants were centrifuged at 800 ϫ g, passed through 0.2-m filters (Nunc, Wiesbaden), and stored at Ϫ80°C.
Treatment of Mice with Parasite Lysates-Mice were treated with 500 l of PBS containing P. chabaudi lysates, heat-inactivated Salmonella typhimurium, or LPS from Salmonella typhosa by intraperitoneal injection. Expression of imap38 in spleens was analyzed either 24 h after a single injection or on day 7 after injections on days 0 and 4.
Reporter Gene Assays-Promoter activity was tested after transient transfection using secreted alkaline phosphatase (SEAP) as reporter gene. The 1872-bp HindIII fragment from plasmid H/H1.87kb was inserted in the 5Ј 3 3Ј direction into the HindIII site of pSEAP2Basic (CLONTECH, Heidelberg). This clone was designated as Ϫ1730/ 146iapSEAP. Vectors pSEAP2Basic and pSEAP2Control were used as negative and positive controls, respectively. This plasmid was digested with XhoI followed by digestion with NheI, PvuII (partial digestion), or Bsp120I. DNA ends were blunted and religated to give rise to Ϫ794/ 146iapSEAP, Ϫ576/146iapSEAP, and Ϫ312/146iapSEAP, respectively. The construct Ϫ1730/96iapSEAP was obtained by digesting Ϫ1730/ 146iapSEAP with EcoRI and religation of the vector. To delete the RSINE1 element from the imap38 promoter, the sequences flanking the SINE were amplified along with the plasmid. 50-l reactions containing 0.3 M of the primers ⌬SINE(Ϫ) 5Ј-AAATACTGATACCCAGTTAATTC-CCCATCC-3Ј and ⌬SINE(ϩ) 5Ј-GGACATGGCTGTTTGCCCCCTAAG-TAAAT-3Ј, 0.2 mM dNTPs, 10 ng of plasmid H/H1.87kb, and 3.75 units of Expand High Fidelity in 1ϫ buffer were denatured for 5 min at 94°C. After 30 cycles 15 s at 94°C, 30 s at 50°C, and 5 min at 72°C, a final extension step at 72°C for 10 min was performed. The parent plasmid was digested by the addition of 10 units of DpnI (Roche Molecular Biochemicals) and incubation at 37°C for 3 h. The PCR product was religated, the resulting plasmid Ϫ1730/146⌬SINEBluescript was digested with HindIII, and the promoter fragment was inserted in pSEAP2Basic, resulting in Ϫ1730/146⌬SINEiapSEAP. Plasmids for transfections were purified by anion exchange chromatography (Qiagen), and endotoxin contaminations were removed on Detoxi-Gel columns (Pierce, St. Augustin, Germany). For assays, cells were seeded in 48-well plates and transfected using 0.1 g of DNA and 0.3 l of FUGENE 6 (Roche Molecular Biochemicals) per well. Transfections were done in triplicate, and each experiment was reproduced 4 -6 times. After 3 days, supernatants were removed, and SEAP activity was determined with the Phospha-Light kit according to the manufacturer's instructions (Tropix, Weiterstadt, Germany). Chemiluminescence was measured in a Lumat LB 9507 luminometer (Berthold, Bad Wildbad, Germany) as relative light units (RLU). Relative promoter activity was calculated as follows: (RLU Ϫ1730/146iapSEAP Ϫ RLU pSEAP2Basic ) ϫ 100/ (RLU pSEAP2Control Ϫ RLU pSEAP2Basic ).
For stimulation experiments, IC-21 and RAW 264.7 cells were transfected with Ϫ1730/146iapSEAP for 24 h. These cells were stimulated by final concentrations of 1 g/ml LPS, 1 ϫ 10 6 lysed P. chabaudi, or 10% spleen cell conditioned medium. After 48 h, culture medium was removed for determination of SEAP and nitrite. Supernatants were centrifuged for 10 min at 14,000 ϫ g and mixed with an equal volume of Griess reagent (1:1 mixture of 2% sulfanilamide (Sigma) in phosphoric acid and 0.2% N-(1-naphtyl)ethylenediamine dihydrochloride (Sigma)). Nitrite concentration was calculated from the A 540 by comparison with an NaNO 2 standard curve in culture medium.

RESULTS
Genomic Organization-Genomic clones containing the complete sequence of the imap38 gene were obtained from C57BL/10 mice by genomic PCR and from 129/ola mice by screening of a cosmid library. In the latter, we identified the two positive clones MPGc121K12460Q2 and MPGc121B23596Q2. A Hin-dIII fragment and a HindIII/ScaI fragment were subcloned from MPGc121K12460Q2 and termed H/H1.87kb and H/Sc2.68kb, respectively. The two latter clones span a region of 4552 bp (Fig. 1A). The genomic PCR fragments RSA, INV, and iapG2 showed 100% identity with the cosmid derived clones over their entire length of 2990 bp. The genomic sequence and the published imap38 cDNA sequence, identical with iap⌬289 in Fig. 1, could be perfectly aligned without interruption, except for an additional cytosine in position 2336 corresponding to position 606 in the original imap38 cDNA sequence (12). This additional cytosine in a very GC-rich context could also be resolved by resequencing the original imap38 cDNA with an automatic sequencer using a high resolution gel matrix. The perfect alignment indicates that the imap38 gene contains no intron.
Transcriptional initiation sites were mapped using high resolution reverse ligation-mediated RACE. 40 PCR-generated clones were randomly picked and sequenced. 14 clones contained specific amplification products, each representing the 5Ј-end of an imap38 mRNA. Six clones began in position 1731, two in position 1790, and six in position 1823. This indicates that the imap38 gene possesses three start sites, with the distal and proximal site being 92 bp apart. Position 1731 is hereby defined as ϩ1. The GC content of the imap38 5Ј-flanking region is 50.5%, which is slightly lower than the overall GC content of 51.4%. In the coding region (between positions 2154 and 2894; see below) the portion of GC base pairs is very high, reaching 63.1%.
Using the program CENSOR, we identified three truncated repetitive elements in the imap38 gene: (i) a RSINE1 element with 74% similarity to its consensus sequence in positions Ϫ131 to Ϫ11 in the proximal promoter; (ii) a retroviral RLTR10A-element with 89% similarity to the consensus in positions ϩ1579 to ϩ1927 bp, surrounding the polyadenylation signal and site of the imap38 gene; (iii) a long interspersed element of the Lx7 family in positions ϩ2246 to ϩ2523 bp with 61% similarity to the consensus.
Chromosomal Assignment-Southern blotting revealed that the imap38 gene is a single copy gene (Fig. 1B). The bands in EcoRI-and TaqI-digested DNA correspond to the size predicted from the genomic sequence, confirming that the cloned intronless DNA sequence represents the only imap38 sequence in the genome. The chromosomal localization of the imap38 gene was determined by fluorescence in situ hybridization to mouse metaphase chromosomes. Microscopic evaluation of the hybridization experiments revealed strong fluorescent signals on mouse chromosome 6. In 93% (27 of 29) of the evaluated metaphase cells, the fluorescent signals were detected on both homologues of chromosome 6 in the distal portion of band B (Fig. 1C). No additional signals in other regions of the mouse genome were observed, thus indicating that the imap38 gene is localized on chromosome 6B.
In Vitro Translation-The longest mature imap38 transcript amounts to 1900 nucleotides and contains two long overlapping ORFs. The first long ORF starts in position 424 at the fourth AUG codon and encodes a polypeptide of 246 amino acids with a molecular mass of 25.8 kDa and a predicted pI of 10.2 (Fig. 2). The other long ORF starts in position 632 at the eighth AUGcodon of the mRNA, coding for a 277-amino acid-long protein with a molecular mass of 31 kDa. The amino acid sequences of both deduced proteins did not show any significant similarity to other known proteins in the Swiss-Prot data base. Start codons of both ORFs are not surrounded by a Kozak sequence for optimal initiation of translation (32). To examine which one of the two long ORFs is used, we subjected two different constructs to in vitro transcription and translation (i.e. the clones iap⌬289 and iapG2). The latter contains the whole 5Ј-UTR starting at position ϩ1, whereas the 5Ј-UTR is truncated in iap⌬289, and AUG 424 is the first start codon. This deletion includes the removal of three short upstream ORFs encoding peptides of 3, 10, and 33 amino acids. In SDS-polyacrylamide gel electrophoresis, the in vitro translation products of iap⌬289 always yielded two bands, a strong band at 26 kDa and a faint band at 23 kDa. The intensity of the latter varied from experiment to experiment (Fig. 3A). Presumably, the 26-kDa product was produced by initiation at AUG 424 , whereas the 23-kDa product was derived from the second or third AUG in the same ORF. Remarkably, the 23-kDa protein became the predominant product, when iapG2 was used as template. Obviously, the 5Ј-UTR containing three short upstream ORFs influenced the initiation of translation. No products of 31 kDa could be detected, suggesting that AUG 632 is not used as an initiation codon at all.
Localization of Tagged IMAP38 -CHO-K1 cells were transiently transfected with expression vectors carrying the imap38 cDNA with most of its 5Ј-flanking region fused to a C-terminal Myc6ϫHis tag. Three different plasmids were used, each carrying the tag in one of the three reading frames. Fig.  3B shows that only pIMAPMycA produced a Myc-tagged protein. In this construct, the tag was fused to the end of the ORF starting at AUG 424 , confirming that only the first open reading frame was used. The protein had the expected molecular mass of 27 kDa. Using confocal laser scanning microscopy, the protein IMAPMycA colocalized with propidium iodide-stained DNA in the nuclei of transfected cells (Fig. 3C). Nuclear localization could also be revealed using an IMAP38 N-terminally tagged with green fluorescent protein (data not shown).
Inducibility in Vivo-Different inbred mouse strains on C57BL/10 or C57BL/6 background expressed imap38 in the spleen at low levels (Fig. 4A). Infections with P. chabaudi induced a dramatic increase in imap38 expression independent on H-2 complex and gender on day 7 postinfection. By contrast, mice with BALB and DBA background, possessing the same H-2 d haplotype as the B10.D2 mice, expressed only very low levels of imap38, even when they were infected with P. chabaudi for 7 days (Fig. 4A). The infection-induced expression of imap38 in B10.A mice was largely restricted to the spleen (Fig. 4B). In comparison with the spleen, there occurred only very low expression in thymus, peritoneal cells, and mesenterial lymph nodes, whereas no expression was detectable in axillary lymph nodes, bone marrow, and peripheral blood, respectively.
When B10.A mice were treated with two doses of parasite lysate derived from 1 ϫ 10 6 P. chabaudi-infected erythrocytes on days 0 and 4, there was also induced expression of imap38 in the spleen on day 7, but to a lesser extent than in P. chabaudi-infected controls (Fig. 5A). However, treatment of control mice with LPS or lysates from blood of noninfected mice did not have any effect on imap38 mRNA level in the spleen. Moreover, a single dose of parasite lysate or heat-inactivated S. typhymurium did not induce imap38 expression (Fig. 5A).
Noninducibility ex Vivo-A broad range of different stimuli including PMA, A23187, anti-CD3, concanavalin A, LPS, IFN-␥, serum from infected mice, spleen cell conditioned medium, and P. chabaudi lysates were tested for their ability to induce imap38 expression in isolated spleen cells, RAW 264.7 and IC-21 macrophage cell lines. Fig. 5B shows that neither LPS, IFN-␥, P. chabaudi lysates, spleen cell conditioned medium alone, nor combinations thereof were able to induce imap38 in cultured spleen cells from B10.A mice. All of the stimuli presented here were able to activate spleen cells as demonstrated by enhanced iNOS transcription. Expression of imap38 could not be elevated by any stimulus tested. Although P. chabaudi lysates were active in vivo, their presence in the culture medium was not sufficient for imap38 induction. We never observed any imap38 expression in the macrophage cell lines.
Promoter Activity-The H/H1.87kb clone contains 1730 bp of the 5Ј-flanking region and 142 bp of the 5Ј-UTR, including all three transcription start sites. Computer analysis of this sequence for cis-acting elements using the program MatInspector in the TRANSFAC data base revealed that the start site (ϩ1) is preceded by a GC-box immediately followed by an E2F-like binding motif (Fig. 6). Such a constellation has been implicated in transcription initiation at several TATA-less promoters (33). The second initiation site (ϩ60) is preceded by two imperfect TATA-box like motifs (AATAAAA). The third site (ϩ93) is surrounded by AA ϩ93 TCTCT, which shows strong homology to the consensus initiator control element sequence (34). Five additional GC-boxes (Sp1 binding sites) and five CAAT-boxes as well as a series of other putative cis-acting elements are outlined in Fig. 6. The close proximity of a 120-bp-long RSINE1 element to the transcriptional initiation site (ϩ1) is rather exceptional.
The activity of the HindIII fragment containing the 1730 bp of the 5Ј-flanking region was compared with the SV40 virus immediate early promoter and enhancer in pSEAP2Control. In the nonmurine CHO-K1 and COS-7 cells, the imap38 promoter exhibited 90 and 42% relative promoter activity, respectively (Fig. 7A). However, less than 10% of the SV40 promoter activity was revealed in macrophage cell lines IC-21 and RAW 264.7 as well as in other mouse cell lines such as fibroblast-like NIH/ 3T3 and L929 cells. Since imap38 is predominantly expressed in splenic macrophages (12), our further analysis of the imap38 promoter concentrated on the two macrophage cell lines. Different stimuli were used to activate the IC-21 and RAW 264.7 cells as confirmed by increased nitrite production. Under these conditions, however, imap38 promoter activity was not upregulated (Fig. 7B). This is in accordance with the above in vitro stimulation experiments, where imap38 induction could not be detected in Northern blots. Moreover, different deletions in the promoter did not result in any activation of promoter activity in the mouse cell lines RAW 264.7 (data not shown) and IC-21 (Fig. 7C). DISCUSSION This study describes cloning of 4552 bp of the malaria-inducible mouse gene imap38, including 1730 bp of the 5Ј-flanking region. The gene does not contain any intron, and the single exon consists of 1900 bp. This feature may lead to consideration of the cloned imap38 gene as a nonfunctional pseudogene generated possibly by a retroposition event. This suspicion would be consistent with the presence of 12-bp-long direct repeats in positions Ϫ298 and ϩ1517, a d(A)-rich sequence (position ϩ1493) between two putative polyadenylation signals (positions ϩ1454 and ϩ1461) and the downstream direct repeat, and the RSINE1-repeat in the proximal promoter. In contrast to this view, however, our data provide evidence that the cloned intronless imap38 gene is functional and its inducibility is obviously subjected to rather complex control mechanisms.
There exists only one single copy of the imap38 gene in the mouse genome that is localized on chromosome 6B. This excludes the possibility that, besides the intronless copy, there is still another intron-containing copy of the imap38 gene present in the mouse genome. In general, intronless genes occur in higher eukaryotes, but they are relatively rare. For instance, only about 6% of the mammalian genes are composed of a single exon (35). Examples of intronless genes are most of the hsp70 genes (36 -39), several G-protein-coupled receptors (40), and genes encoding nuclear proteins such as all histones (41), some protamines (42,43), and transcription factors such as C/EBP␤ (44), XLPOU3 (45), POUF3 (46), BRN-3b (47), Pw1 (48), c-Mos (49), and Sry (50,51).
The functionality of the imap38 gene is further supported by the fact that the imap38 mRNA is translatable both in vitro and in transfected cells. In particular, our data show the usage of the long ORF beginning at AUG 424 encoding a highly basic 25.8-kDa protein as confirmed by in vitro translation and Western blot analysis of CHO-K1 cells transiently transfected with an imap38 cDNA construct containing a C-terminal Myc6 ϫ His-tag. In vitro translation using an imap38 cDNA with a truncated 5Ј-UTR resulted in production of a 26-kDa protein and trace amounts of a 23-kDa protein. However, when a template with the complete 5Ј-UTR was used, the 23-kDa product was produced more efficiently than the 26-kDa protein. This suggests that the structure of the 5Ј-UTR is involved in regulation of imap38 mRNA translation (52,53). Both the 26-kDa protein and the 23-kDa protein are deduced to be highly basic. The IMAP38 protein is localized in nuclei, but not in nucleoli of CHO-K1 cells transiently transfected with either a Myc6ϫHis-tagged imap38 construct or a GFP-imap38 construct. This nuclear localization suggests that IMAP38 is not a direct anti-parasite effector molecule but rather is a nuclear regulatory protein in cells responding to P. chabaudi infections.
The complex mechanisms controlling inducibility of imap38 become evident at several different levels. Our data demonstrate that inducibility of imap38 in spleens is under control of genes of the non-H-2 background, whereas genes of the H-2 complex and gender have no influence. These non-H-2 genes are present in C57BL/10-and C57BL/6-derived mouse strains and allow expression of imap38, while imap38 cannot be induced in BALB/c and DBA2/J mice. Moreover, imap38 expression in vivo is extremely tissue-and cell-specific; it is almost exclusively restricted to the spleen at maximal parasitemia, i.e. on day 7 after infections with P. chabaudi. At this time, other lymphoid organs such as blood, bone marrow, thymus, and lymph nodes are also activated and involved in the defense against P. chabaudi infections. Remarkably, however, these organs express imap38 only in traces if at all. Also, imap38 is barely detectable in nonlymphoid organs such as brain, heart, kidney, and liver, as shown previously (12). In the spleen, imap38 expression is confined to cells of the adherent fraction, i.e. predominantly macrophages, and Ig ϩ B-cells, whereas it is only scarcely detectable in Ig Ϫ T-cells if at all (12). By contrast, macrophages from the peritoneum as well as macrophages and B-cells in lymph nodes do not express imap38, although these cells also come in close contact with P. chabaudi components during an infection. These data indicate that the cell-specific induction of the imap38 gene requires obviously specific spleen factor(s). Such factors are provided only in the intact organ and work synergistically with parasite components. This can be concluded from our data indicating that lysed parasites can also induce, although to a lesser extent than P. chabaudi infections, expression of imap38 in the spleen but not in spleen cells ex vivo. Activation of macrophages and B-cells per se is not sufficient for imap38 induction, since both LPS and parasite lysates do activate macrophages, as indicated by increased levels of iNOS mRNA ex vivo, but only P. chabaudi lysates show a positive effect on imap38 expression in vivo.
Integration of both parasite and spleen signals is expected to occur at the promoter of imap38. We have cloned 1730 bp of the imap38 5Ј-flanking region. This promoter exhibits an unusual structure with three transcription initiation sites and a SINE element adjacent to the first start site. There is information available that repetitive elements can have strong positive or negative effects on transcription of adjacent genes. In particular, cis-acting SINEs or partial SINEs in the 5Ј-flanking regions or in introns have been shown to carry transcription factor binding sites (54,55) and to function either as enhancers (56,57) or repressors of transcription (58 -60), but none of these repeats has been reported to be located as close to a transcription start site as RSINE1 in the imap38 gene. In addition, there are also numerous transcription factor binding sites such as 6 GC-boxes, 5 CAAT-boxes, and several binding sites for LPS-or cytokine-inducible transcription factors, i.e. NF-B, C/EBP␤, AP-1, Pu.1, Ets, Elk, STAT, and IRF-1. Reporter gene assays revealed that the promoter was active in nonmurine cell lines but showed only marginal activity in mouse macrophage cell lines such as IC-21 and RAW 264.7 as well as in mouse NIH/3T3 and L929 fibroblasts. This minor activity is in correspondence with the stringent expression pattern in vivo, i.e. restriction of expression to splenic macrophages and B-cells. Moreover, promoter activity was not enhanced in deletion constructs in vitro, suggesting that stringent control is a feature of the proximal promoter itself and not mediated by a distal silencer. This repression of promoter activity may also be important for tissue-specific control in vivo. Stimulation of transfected IC-21 and RAW 264.7 macrophages with LPS, P. chabaudi lysates, and spleen cell conditioned medium resulted in increased production of NO 2 Ϫ but did not enhance promoter activity as expected from the above mentioned stimulation experiments with these cell lines and cultured spleen cells. Unfortunately, cell lines derived from spleens of C57BL/10 or C57BL/6 mice are not available.
Finally, there is a conspicuous correlation between imap38 FIG. 7. Analysis of the imap38 promoter by SEAP reporter gene assay. A, the relative activity of the imap38 promoter was tested in different cell lines. The Ϫ1730/142iapSEAP construct was compared with the strong SV40 immediate early promoter and enhancer present in pSEAP2Control. B, activity of the imap38 promoter in stimulated IC-21 macrophages. Cultures were transfected with Ϫ1730/146iapSEAP and stimulated as indicated. Nitrite concentration in the culture medium was determined as a control of macrophage activation. C, deletion analysis of imap38 promoter in IC-21 macrophages. Cells were transfected with the indicated deletion constructs, and promoter activity was determined in relation to pSEAP2Control. Data represent the mean and half S.D. of 4 -6 independent experiments. inducibility and the ability to develop protective immunity to P. chabaudi malaria in mice with C57BL/10 and C57BL/6 non-H-2 background. By contrast, P. chabaudi infections do not induce imap38 and have a lethal outcome in mice on BALB and DBA background. Obviously, imap38 expression is regulated by genes of that non-H-2 background which also controls resistance to P. chabaudi blood stage malaria. Moreover, the inducibility of imap38 in macrophages and B-cells during acute malaria coincides with a dramatic remodeling of spleen architecture, resulting in dramatic spleen enlargement and improved filter capacity for removing P. chabaudi-infected erythrocytes (4,5,61,62). This, together with the fact that imap38 remains constitutively expressed at a high level in immune mice for at least 13 weeks postinfection (12), allows us to envisage the imap38 gene as critically involved in spleen reorganization, providing the structural basis for immune mechanisms to become effective against parasites.