Supplementary material for Borkhardt et al., Proc. Natl. Acad. Sci. USA, 10.1073/pnas.150079597
Materials and Methods
Competitive Semiquantitative Reverse Transcription–PCR Analysis, GENESCAN Analysis
. cDNA was prepared from lymphocytes of healthy volunteers and leukemic cells of patient 9 as described above. Two microliters of cDNA was subjected to 25 cycles of PCR at 94°C at 15 sec, 60°C for 30 sec, and 72°C for 45 sec with the help of primers I and IV. They had the same primer template sequence for both inserted vs. noninserted cDNA (see Fig. 7B.). Primer I carried a fluorescein label at its 5' end to allow further analysis by a GENESCAN procedure. For this purpose, 1 µl of the final PCR product was diluted with 9 µl sterile water. A master mix corresponding to 0.5 µl GENESCAN standard (GENESCAN 2500TMRox, Perkin–Elmer) and 2.5 µl of formamide per sample was prepared. One microliter of the diluted PCR product was added to 3 µl of this master mix. The samples were heated to 90°C for 3 min and quenched on ice. Four microliters of each sample was subjected to electrophoresis over a 6% polyacrylamide gel in the automated DNA sequencer 373A (Perkin–Elmer). Gels were analyzed on an Apple McIntosh IIIci computer using the GENESCAN software as supplied by the manufacturer. The specific amplification of the cDNA fragment with insertion in healthy volunteers was done by nested PCR with primers I and IV (first round) and II and III (second round). Amplification was carried out for 35 cycles with denaturation at 94°C for 10 sec, annealing at 55°C for 30 sec, and strand extension for 30 sec at 72°C. In the second round of PCR the number of cycles was reduced to 25.Fluorescence in Situ Hybridization (FISH)
. In the patient with the t(5;11), double-color FISH was performed as described with a painting probe specific for chromosome 5 and a 440-kb long yeast artificial chromosome clone, 13 HH4, that previously had been shown to span the breakpoints of various translocations involving this gene region. The painting probe was labeled with biotin-21-dUTP (CLONTECH) by DOP-PCR and detected with FITC. The yeast artificial chromosome clone was labeled with digoxigenin-11-dUTP and detected with sheep anti-digoxigenin rhodamine (Boehringer Mannheim). For mapping onto normal metaphases, the 1,167-bp long clone of the human Graf gene was labeled in an identical fashion. The slides were counterstained with 4,6-diamino-2-phenylindole (DAPI) (Sigma), air dried, and embedded with H-1000 mounting medium (Vectashield, Vector Laboratories). The slides were analyzed with a Zeiss-Axioplan fluorescence microscope and an image analysis system with the IP LAB 3.0 software (Vysis, Downers Grove, IL).Patients and Cell Lines
. Cryopreserved samples of mononuclear bone marrow cells from 73 patients with myelodysplastic snydrome (MDS) and acute myeloid leukemia (AML), and a deletion (5q) were obtained from the Department of Internal Medicine I, University of Vienna, Austria, and the Pediatric Department, University of Giessen, Germany. The cell lines were purchased from Hans Drexler, Deutsche Sammlung für Mikroorganismen und Zellkulturen, Braunschweig, Germany.Southern, Northern, Multiple Tissue, and Zoo Blot
. For analysis of the MLL gene, high molecular weight DNA was isolated from the sample with the t(5;11) according to standard procedures. Five micrograms of DNA was digested with BamHI (Boehringer Mannheim). The DNA was separated by electrophoresis in a 0.7% agarose gel at 25 V overnight and hybridized onto nylon membranes. The blot were probed with a 859-bp BamHI fragment of cDNA that spans the MLL-breakpoint cluster region defined by exons 5-11 of the MLL gene. The prefabricated Northern blots and the Zoo blot were purchased from CLONTECH. The blots were hybridized according to the manufacturer’s instructions with a cDNA clone of the human Graf gene ranging from nucleotides 225 to 1,392. The cDNA probe was labeled by incorporating digoxigenine 11-dUTP. Filters were hybridized at 37°C for 10 h by using the EasyHyb solution (Boehringer Mannheim), followed by two washes with 2´ SSC/0.1% SDS and 0.1´ SSC/0.1% SDS, respectively. The detection was carried out with an anti-digoxigenin antibody (Boehringer Mannheim) that was conjugated with alkaline phosphatase.Isolation of the P1 Clone Containing Graf. The human P1 library was obtained from the German Resource Center for the Human Genome Project, Berlin. The 22 ´ 22-cm filters were hybridized with 330 ng of the Graf cDNA clone. The probe was dissolved in 13.2 ml of EasyHyb solution, achieving a final concentration of 25 ng/ml. The filters were hybridized for 16 h at 42°C, followed by the washes above described for Southern blotting. The detection was done at room temperature for 30 min with 13.2 µl of anti-digoxigenin antibody dissolved in 132 ml of blocking buffer (Boehringer Mannheim). After washing, filters were covered with 4 ml of CSPD [disodium 3-(4-methoxyspirotricyclodeanphenyl phosphate)] solution (Boehringer Mannheim) for 5 min and exposed between two intensifying screens for 40 min.
Immunohistochemistry
. Normal human tissues were taken from biopsies after surgery or were obtained less than 24 h postmortem at the Department of Pathology, Justus-Liebig-University (Giessen, Germany). Patients and deceased persons (n = 122) were between 4 months and 86 years old. The different tissues were either snap-frozen in liquid nitrogen, embedded in OCT compound (Miles) and stored at –80°C until use or fixed in 4.5% formalin, pH 7.0 for at least 24 h, routinely processed, and embedded in paraffin.Two synthetic peptides consisting of amino acids 659–672 (subsequently called ED98015) or 677–692 (ED98016) of Graf were used as antigens for the production of polyclonal antisera. The synthesis of peptide antigens, immunization of rabbits, and collecting of antisera was commercially done by Eurogentec, Brussels. Preimmune serum of each rabbit was used as negative control.
A stable KMST6 cell line was transfected with the expression plasmid pcDNA3.1 (Invitrogen) containing the entire coding region of GRAF and subsequently used as a positive control to verify the specificity of antisera ED98015 and ED98016. For transfections, KMST6 cells were plated at 2 ´ 105 cells per ml and transfected with the Superfect reagent (GIBCO) using 10 µg of plasmid DNA per 100 cm2 dish. Cells were incubated with the transfection mixture for 6 h, subsequently rinsed, and incubated for additional 48 h before harvesting. Transfected and untransfected KMST6 cells were separately grown in Falcon culture slides and stimulated with PMA and ionomycin 4 h before staining. The cells were fixed in acetone followed by incubation with the antisera ED98015 or ED98016 (1:1,000) or with preimmune serum (1:1,000) for 30 min in a humid chamber.
Immunohistochemistry on Human Tissue Specimens
. Cryostat or microtome sections (5 µm) of the specimens were mounted on slides and air-dried, and frozen sections were stored at –20°C until immunohistochemistry was performed. Hematoxylin/eosin-stained sections served to rule out autolytic tissue damage in postmortem tissues. Immunohistochemical detection of GRAF protein was performed as described by the highly sensitive alkaline phosphatase (APAAP) technique. Frozen sections were fixed in acetone for 10 min at room temperature, and paraffin sections were dewaxed for 10 min in xylol, followed by 10 min acetone and 10 min 50% acetone/50% TBS, pH 7.4. For the detection of GRAF protein, paraffin sections were microwave-treated by boiling three times for 5 min at 600 W in 0.01 M citrate buffer, pH 6.0. Frozen sections and paraffin sections were incubated for 30 min with the GRAF antisera ED98015 and ED98016, 1:1,000 diluted for frozen sections and 1:200 for paraffin sections. Three subsequent detection steps followed, each lasting 30 min: an incubation with mouse anti-rabbit Ig (Dako, 1:500 in RPMI 1640 medium, Life Technologies, Paisley, U.K., supplemented with 12.5% inactivated bovine serum to inhibit unspecific crossreactivity), an incubation with rabbit anti-mouse Ig (Dako, rabbit-"link", 1:40 in RPMI 1640 medium supplemented with 12.5% inactivated bovine serum), and an incubation with the APAAP complex (1:50, Dako). For paraffin sections the link and APAAP steps were repeated for 10 min. Control sections were incubated with ED98015 and ED98016 preimmune serum (1:1,000) and mouse anti-rabbit Ig (clone: MR12/52, 1 µg/ml, Dako) as primary antibodies. Immunostaining of insulin and glucagon was performed on paraffin sections by using a monoclonal mouse anti-insulin antibody (1:200, clone: E2E3, Immunotech, Marseille, France, 30 min) and a polyclonal rabbit anti-swine glucagon antiserum with well-known cross reactivity against human glucagon (1:500, A0565, Dako, 30 min). For insulin detection the link and APAAP steps were applied, The samples were thoroughly washed in TBS, pH 7.4, between the steps. Alkaline phosphatase substrate reaction with new fuchsin (100 µg/ml) and levamisole (400 µg/ml) was performed for 20 min at room temperature. For glucagon detection immunohistochemistry was performed with biotinylated swine anti-rabbit Ig (1:300, Dako, 30 min), streptavidin/HRP P397 (1:300, Dako, 30 min), and 4-min incubation in DAB chromogen solution (Dako). Sequential immunohistochemical double staining was used as described [Bohle, R. M., Brettreich, S., Repp, R., Borkhardt, A., Kosmehl, H. & Altmannsberger, H. M. (1996) Am. J. Pathol. 148, 737-–738) with slight modifications: 3 ´ 5 min microwave pretreatment, anti-insulin antibody 1:50 with LSAB detection (Dako ChemMate Detection Kit Peroxidase/ DAB rabbit/mouse) according to the manufacturer’s protocol with an additional 15 min 3% H2O2 incubation after the streptavidin peroxidase step, GRAF antiserum ED98015 1:750 with mouse anti-rabbit Ig (1:500, Dako, 30 min), and APAAP detection (Dako ChemMate Detection Kit APAAP mouse) according to the manufacturer’s protocol. Sections were counterstained with hematoxylin and mounted in gelatin.Results
Immunohistochemical Findings
. Comparison of the immunostaining patterns of freshly fixed cryostat tissue sections with the respective formalin-fixed, paraffin-embedded material by two independent investigators (R.M.B. and P.K.) confirmed that the distribution and intensity of labeling was comparable.Epithelial Cells
. The GRAF protein was abundantly expressed in the cytoplasm of the vast majority of epithelial cells, irrespective of their origin, such as in those of the squamous (and superficial) skin (Fig. 4G), esophagus, cervix, urothelium, eccrine/serous glandulae of the respiratory system, upper gastrointestinal tract, superficial colon, ductal and acinar part of mammary gland, and reproductive system. Hepatocytes showed a weak to moderate cytoplasmatic staining. In some instances only pericentral hepatocytes but not periportal hepatocytes were positive. The respiratory and bile duct epithelium as well as the proximal and distal tubules of the kidney were weakly labeled. The cytoplasm of the mucous epithelia of the upper gastrointestinal tract and respiratory system, sebaceous glands of the skin, basal cells of squamous epithelium, colon crypts, and myoepithelial cells were negative (Table 2). In addition to this cytoplasmatic distribution, nuclear staining was seen in scattered cells in squamous epithelia of the cervix and esophagus, ductal and acinar epithelia of the mammary gland (Fig. 4 D and E), sweat glands, respiratory epithelium, urothelium, fallopian tube epithelium, certain serous epithelia, and testis germ cells.Endocrine Tissues
. In endocrine tissues, the GRAF protein was expressed heterogeneously. An only weak cytoplasmic staining was observed in the zona glomerulosa and reticularis of the adrenal gland, whereas the zona fasciculata and medullary cells were negative. The follicular epithelium of the thyroid was weakly labeled. Leydig cells showed a stronger cytoplasmatic staining. However, we found a striking cytoplasmatic labeling of many islets of Langerhans cells (Fig. 4K). Single and double staining with an anti-insulin antibody identified these cells as the ß cells of the pancreas (Fig. 4 L and M), whereas glucagon-producing cells were GRAF-negative (data not shown).Lymphatic and Hematopoeitic System
. The lymphocytes of follicules as well as the T cell zones of lymph nodes, as well as the spleen and tonsil were negative (Fig. 4I). In the bone marrow, the cytoplasm of normoblasts were clearly labeled and that of the megakaryocytes weakly labeled, whereas, somewhat surprisingly, the GRAF protein was absent in all myelopoetic cells.Other Mesenchymal Tissues
. Smooth muscle and peripheral nerve cells contained only little or no GRAF protein, but ganglion cells from enteric ganglia were moderately stained (Fig. 4H). Striated and cardiac muscle as well as brown adipose tissue was weakly positive. For details of the GRAF protein expression see Table 2.