Large-scale purification of hnRNP proteins from HeLa cells by affinity chromatography on ssDNA-cellulose

Measurement of the frequency of histone displacement during the in vitro transcription of nucleosomes: RNA is a competitor for these histones

Transcription through tandemly arranged nucleosomes was studied to determine the frequency at which the nucleosomes would disrupt and cause displacement of the associated histones to a competitor DNA. In order to more effectively preserve topological effects, the template that was used in the in vitro transcription system was a large covalently, closed circular plasmid (8.9 kb). The plasmid contained two promoters for T7 RNA polymerase, each separated by 4.4 kb, and transcription was done in the presence of topoisomerase I at physiological ionic strength. Nucleosome disruption was observed at an approximate frequency of 1 in 4 nucleosomes such that after several rounds of transcription on the plasmid 80% of the nucleosomes were disrupted. Unexpectedly, all four histones were found associated with the RNA rather than the competitor DNA. The histones bound the competitor DNA only after removal of the RNA by RNase A treatment. By analyzing the topological state of the competitor DNA, it was observed that the majority of the histones that were displaced from the RNA were able to re-form nucleosomes. Additional experiments were done to determine the reasons for the preferential binding of histones to the newly synthesized RNA. It was found that the large molecular weight RNA binds histones with an approximate 100-fold greater affinity relative to DNA when at physiological ionic strength. Within the cell, this high-affinity binding would be expected to require cellular mechanisms to regulate the interaction of RNA with histones. The relatively high frequency of displacement of all four histones during transcription is higher than what is observed in vivo and suggests that additional factors are needed to regulate this displacement. These observations are discussed and compared with previous studies that have examined the process of transcription through nucleosomes.

Identification of autoantibodies to the I protein of the heterogeneous nuclear ribonucleoprotein complex in patients with systemic sclerosis

OBJECTIVE

To assess the presence of autoantibodies to the 1 protein (polypyrimidine-tract binding protein) of the heterogeneous nuclear RNPs (hnRNP) in different connective tissue diseases. Antibodies to other hnRNP proteins (A1, A2, and B) have been previously found in patients with rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), and mixed connective tissue disease (MCTD).

METHODS

Sera from 101 patients with various connective tissue diseases and 25 normal controls were investigated by enzyme-linked immunosorbent assay and immunoblotting, for their reactivity to highly purified recombinant hnRNP I. Moreover, reactivity to cellular hnRNP I protein was investigated by immunoblotting using a partially purified preparation of hnRNP proteins (including A1, A2, B, and I), and by indirect immunofluorescence. For the analysis of the fluorescence pattern, affinity-purified antibodies to hnRNP I; obtained from a selected patient, were tested on HEp-2 cells.

RESULTS

By immunoblotting, antibodies reacting to recombinant hnRNP I were found in 22 of 40 patients with systemic sclerosis (SSc), 3 of 32 with RA, 0 of 23 with SLE, and 0 of 6 with MCTD. Antibodies to recombinant hnRNP I were more frequently found in patients with pre-SSc or limited SSc (15 of 24) than in those with intermediate or diffuse SSc (7 of 16). In indirect immunofluorescence studies, affinity-purified anti-hnRNP I autoantibodies gave a diffuse nucleoplasmic staining. Using an hnRNP preparation from nuclear extracts, anti-hnRNP I reactivity was detectable in SSc sera, while it was not detectable in RA, SLE, and MCTD sera reacting with hnRNP A/B proteins.

CONCLUSIONS

Human autoimmune sera show distinct patterns of anti-hnRNP reactivity, i.e., anti-A/B in SLE and RA sera, and anti-I in SSc sera. This suggests that A/B proteins and the I protein may be involved in different dynamic hnRNP complexes that elicit different autoimmune responses. From a clinical perspective, anti-hnRNP I antibodies are frequently associated with pre-SSc features, suggesting an early appearance of these antibodies during the course of the disease.

The roles of heterogeneous nuclear ribonucleoproteins (hnRNP) in RNA metabolism

In eukaryotic cells, messenger RNAs are formed by extensive post-transcriptional processing of primary transcripts, assembled with a large number of proteins and processing factors in ribonucleoprotein complexes. The protein moiety of these complexes mainly constitutes a class of about 20 major polypeptides called heterogeneous nuclear ribonucleoproteins or hnRNPs. The function and the mechanism of action of hnRNPs is still not fully understood, but the identification of RNA binding domains and RNA binding specificities, and the development of new functional assays, has stimulated interest in them. In contrast to previous models that hypothesised a mere structural (histone-like) function, a more diversified and dynamic role for these proteins is now emerging. In fact, they can be viewed as a subset of the trans-acting pre-mRNA maturation factors. They might actively participate in post-transcriptional events such as regulated splicing and mRNA export. Moreover, recent data suggest an involvement of some of these proteins in molecular diseases. Here we present an overview of the most relevant properties of hnRNPs and discuss some emerging ideas on their roles.

RNP1, a new ribonucleoprotein gene of the yeast Saccharomyces cerevisiae

A previously unidentified ribonucleoprotein (RNP) gene of yeast has been cloned and sequenced. The gene, named RNP1, was found adjacent to a previously sequenced gene encoding the second gene for ribosomal protein L4. RNP1 contains two RNA Recognition Motifs (RRM), [alternatively known as RNA binding Domains (RBD)], but unlike most RNP genes does not contain any auxiliary simple sequence domains. The first RRM (RRM1) most resembles RRM domains found in the hnRNP A/B class of RNP proteins. The second RRM (RRM2) most resembles a RRM so far seen only in the single RRM of the yeast SSB1 gene. Two null mutants of RNP1 that were created, a frameshift disruption and a complete deletion of the gene, were viable, demonstrating that the gene is not essential for cell growth. Two double null mutants of yeast RNP genes that were created (delta RNP1/delta SSB1 and delta SSB1/delta NPL3) were also viable. A fragment identical in size to the RRM1 domain could be amplified by PCR from the DNA of fungi, plants, and animals, using primers matching the ends of this domain, indicating that the structure of RRM1 is conserved. Another potential open reading frame on the same cloned fragment of DNA encodes a gene product whose structure resembles that of a seven-transmembrane-segment membrane receptor protein.

The existence of eukaryotic ribonucleoprotein consensus sequence-type RNA-binding proteins in a prokaryote, Synechococcus 6301

A group of proteins containing a conserved ribonucleoprotein consensus sequence (RNP-CS)-type RNA-binding domain (CS-RBD) of approximately 80 amino acids is present in eukaryotic cells and binds specifically to a wide variety of RNA molecules. We have isolated 12 kDa single-stranded DNA binding proteins from the unicellular cyanobacterium Synechococcus 6301. The amino-terminal sequence was determined and two distinct genomic clones were isolated from a Synechococcus 6301 genomic library. Sequence analysis revealed that two closely related proteins contain a single CS-RBD of 82 amino acids and are named as 12RNP1 and 12RNP2. Both of the CS-RBDs share the highest amino acid identity with those of chloroplast ribonucleoproteins (40-51%). The 12RNP proteins were expressed in Escherichia coli bearing plasmids encoding glutathione S-transferase/12RNP fusion proteins and subjected to in vitro nucleic acid-binding assay. Both 12RNP1 and 12RNP2 bind to RNA homopolymers poly(U) and poly(G), indicating that they might be RNA-binding proteins. This is the first example of such proteins in prokaryotes. The 12RNP1 and 12RNP2 genes are transcribed as monocistronic mRNAs and the steady-state mRNA level of 12RNP1 is over 20-fold than that of 12RNP2. Due to the easiness of genetic manipulations the cyanobacterium will provide an excellent system to analyze the function of not only cyanobacterial but also plant RNA-binding proteins.

Purification of ribonucleoproteins by a novel approach: isolation of the SSB1 ribonucleoprotein from yeast and demonstration that it has no role in mRNA splicing

A novel approach is described to purify potential ribonucleoproteins (RNP) of yeast. The method assays a yeast RNP complex, assembled in vitro on actin pre-mRNA, by low-ionic strength acrylamide gel electrophoresis. The minimal protein components of this RNP complex were three proteins, one of 30 kDa and two at 42-44 kDa, defined by formation of the complex on biotinylated-RNA, binding of this complex to avidin-agarose, and salt elution of the protein in the biotinylated-RNP complex. Using the assay for RNP complex formation, an RNP protein was purified to homogeneity on the basis of its affinity towards single-stranded DNA and RNA. This RNP protein turned out to be identical to a known RNP protein, the single-stranded binding protein 1 (ssb1) of yeast, on the basis of identical gel electrophoretic migration, antibody cross-reactivity, and identical properties on the gel complex formation assay. In vitro mRNA splicing was normal in extracts made from a yeast strain missing ssb1 (ssb1- strain). Addition of anti-ssb1 antibody to splicing extracts made from a wild type strain did not inhibit or diminish splicing. Instead, mRNA splicing was reproducibly stimulated several fold, indicating competition between ssb1 and splicing factors for binding to single-stranded RNA in the extracts. RNP complexes still formed in the ssb1- strain, demonstrating that it would be possible to purify other RNP proteins from this strain using the gel complex formation assay.

Differential binding of heterogeneous nuclear ribonucleoproteins to mRNA precursors prior to spliceosome assembly in vitro

We have investigated the composition of the earliest detectable complex (H) assembled on pre-mRNA during the in vitro splicing reaction. We show that most of the proteins in this complex correspond to heterogeneous nuclear ribonucleoproteins (hnRNP), a set of abundant RNA-binding proteins that bind nascent RNA polymerase II transcripts in vivo. Thus, these studies establish a direct parallel between the initial events of RNA processing in vitro and in vivo. In contrast to previous studies, in which total hnRNP particles were isolated from mammalian nuclei, we determined the hnRNP composition of complexes assembled on individual RNAs of defined sequence. We found that a unique combination of hnRNP proteins is associated with each RNA. Thus, our data provide direct evidence for transcript-dependent assembly of pre-mRNA in hnRNP complexes. The observation that pre-mRNA is differentially bound by hnRNP proteins prior to spliceosome assembly suggests the possibility that RNA packaging could play a central role in the mechanism of splice site selection, as well as other posttranscriptional events.

hnRNP I, the polypyrimidine tract-binding protein: distinct nuclear localization and association with hnRNAs

Many hnRNP proteins and snRNPs interact with hnRNA in the nucleus of eukaryotic cells and affect the fate of hnRNA and its processing into mRNA. There are at least 20 abundant proteins in vertebrate cell hnRNP complexes and their structure and arrangement on specific hnRNAs is likely to be important for the processing of pre-mRNAs. hnRNP I, a basic protein of ca. 58,000 daltons by SDS-PAGE, is one of the abundant hnRNA-binding proteins. Monoclonal antibodies to hnRNP I were produced and full length cDNA clones for hnRNP I were isolated and sequenced. The sequence of hnRNP I (59,632 daltons and pI 9.86) demonstrates that it is identical to the previously described polypyrimidine tract-binding protein (PTB) and shows that it is highly related to hnRNP L. The sequences of these two proteins, I and L, define a new family of hnRNP proteins within the large superfamily of the RNP consensus RNA-binding proteins. Here we describe experiments which reveal new and unique properties on the association of hnRNP I/PTB with hnRNP complexes and on its cellular localization. Micrococcal nuclease digestions show that hnRNP I, along with hnRNP S and P, is released from hnRNP complexes by nuclease digestion more readily than most other hnRNP proteins. This nuclease hypersensitivity suggests that hnRNP I is bound to hnRNA regions that are particularly exposed in the complexes. Immunofluorescence microscopy shows that hnRNP I is found in the nucleoplasm but in addition high concentrations are detected in a discrete perinucleolar structure. Thus, the PTB is one of the major proteins that bind pre-mRNAs; it is bound to nuclease-hypersensitive regions of the hnRNA-protein complexes and shows a novel pattern of nuclear localization.

Differential binding of heterogeneous nuclear ribonucleoproteins to mRNA precursors prior to spliceosome assembly in vitro

We have investigated the composition of the earliest detectable complex (H) assembled on pre-mRNA during the in vitro splicing reaction. We show that most of the proteins in this complex correspond to heterogeneous nuclear ribonucleoproteins (hnRNP), a set of abundant RNA-binding proteins that bind nascent RNA polymerase II transcripts in vivo. Thus, these studies establish a direct parallel between the initial events of RNA processing in vitro and in vivo. In contrast to previous studies, in which total hnRNP particles were isolated from mammalian nuclei, we determined the hnRNP composition of complexes assembled on individual RNAs of defined sequence. We found that a unique combination of hnRNP proteins is associated with each RNA. Thus, our data provide direct evidence for transcript-dependent assembly of pre-mRNA in hnRNP complexes. The observation that pre-mRNA is differentially bound by hnRNP proteins prior to spliceosome assembly suggests the possibility that RNA packaging could play a central role in the mechanism of splice site selection, as well as other posttranscriptional events.

Characterization and primary structure of the poly(C)-binding heterogeneous nuclear ribonucleoprotein complex K protein

At least 20 major proteins make up the ribonucleoprotein (RNP) complexes of heterogeneous nuclear RNA (hnRNA) in mammalian cells. Many of these proteins have distinct RNA-binding specificities. The abundant, acidic heterogeneous nuclear RNP (hnRNP) K and J proteins (66 and 64 kDa, respectively, by sodium dodecyl sulfate-polyacrylamide gel electrophoresis) are unique among the hnRNP proteins in their binding preference: they bind tenaciously to poly(C), and they are the major oligo(C)- and poly(C)-binding proteins in human HeLa cells. We purified K and J from HeLa cells by affinity chromatography and produced monoclonal antibodies to them. K and J are immunologically related and conserved among various vertebrates. Immunofluorescence microscopy with antibodies shows that K and J are located in the nucleoplasm. cDNA clones for K were isolated, and their sequences were determined. The predicted amino acid sequence of K does not contain an RNP consensus sequence found in many characterized hnRNP proteins and shows no extensive homology to sequences of any known proteins. The K protein contains two internal repeats not found in other known proteins, as well as GlyArgGlyGly and GlyArgGlyGlyPhe sequences, which occur frequently in many RNA-binding proteins. Overall, K represents a novel type of hnRNA-binding protein. It is likely that K and J play a role in the nuclear metabolism of hnRNAs, particularly for pre-mRNAs that contain cytidine-rich sequences.

Isolation of hnRNP complexes from Drosophila melanogaster

Nascent RNA polymerase II transcripts, heterogeneous nuclear RNAs (hnRNAs), become associated with nuclear proteins (hnRNP Proteins), and their processing into mRNAs takes place in these hnRNP complexes. hnRNP complexes have previously been purified from vertebrate cells. Here we report the isolation of hnRNP complexes from an invertebrate organism, the fruitfly Drosophila melanogaster. Candidate hnRNP proteins were purified from D. melanogaster embryos by ssDNA affinity chromatography, and mAbs were produced to many of the major proteins. Genuine hnRNP proteins were identified by several criteria, including nucleoplasmic localization, association with nascent transcripts, crosslinking to poly(A)-containing RNA in living cells, and amino acid sequence. In addition, mAbs that cross-react between the fruitfly and human hnRNP proteins were obtained. Most importantly, using hnRNP-specific mAbs we have purified the hnRNP complexes from D. melanogaster cells. These RNAase-sensitive complexes contain at least 10 major proteins designated hrps, the most abundant proteins having apparent molecular masses of 36, 38, 39, 40, 44, 48, 54, 62, 70, and 75 kD. cDNAs and complete sequences for several of these proteins have been obtained and are presented in the accompanying paper (Matunis, E. L., M. J. Matunis, and G. Dreyfuss. 1992. J. Cell Biol. 116:257-269). The purification of D. melanogaster hnRNP complexes will facilitate genetic and cytological studies on the function of hnRNA-binding proteins and on the posttranscriptional regulation of gene expression.

Characterization and primary structure of the poly(C)-binding heterogeneous nuclear ribonucleoprotein complex K protein

At least 20 major proteins make up the ribonucleoprotein (RNP) complexes of heterogeneous nuclear RNA (hnRNA) in mammalian cells. Many of these proteins have distinct RNA-binding specificities. The abundant, acidic heterogeneous nuclear RNP (hnRNP) K and J proteins (66 and 64 kDa, respectively, by sodium dodecyl sulfate-polyacrylamide gel electrophoresis) are unique among the hnRNP proteins in their binding preference: they bind tenaciously to poly(C), and they are the major oligo(C)- and poly(C)-binding proteins in human HeLa cells. We purified K and J from HeLa cells by affinity chromatography and produced monoclonal antibodies to them. K and J are immunologically related and conserved among various vertebrates. Immunofluorescence microscopy with antibodies shows that K and J are located in the nucleoplasm. cDNA clones for K were isolated, and their sequences were determined. The predicted amino acid sequence of K does not contain an RNP consensus sequence found in many characterized hnRNP proteins and shows no extensive homology to sequences of any known proteins. The K protein contains two internal repeats not found in other known proteins, as well as GlyArgGlyGly and GlyArgGlyGlyPhe sequences, which occur frequently in many RNA-binding proteins. Overall, K represents a novel type of hnRNA-binding protein. It is likely that K and J play a role in the nuclear metabolism of hnRNAs, particularly for pre-mRNAs that contain cytidine-rich sequences.

Diversity of a ribonucleoprotein family in tobacco chloroplasts: two new chloroplast ribonucleoproteins and a phylogenetic tree of ten chloroplast RNA-binding domains

Two new ribonucleoproteins (RNPs) have been identified from a tobacco chloroplast lysate. These two proteins (cp29A and cp29B) are nuclear-encoded and have a less affinity to single-stranded DNA as compared with three other chloroplast RNPs (cp28, cp31 and cp33) previously isolated. DNA sequencing revealed that both contain two consensus sequence-type homologous RNA-binding domains (CS-RBDs) and a very acidic amino-terminal domain but shorter than that of cp28, cp31 and cp33. Comparison of cp29A and cp29B showed a 19 amino acid insertion in the region separating the two CS-RBDs in cp29B. This insertion results in three tandem repeats of a glycine-rich sequence of 10 amino acids, which is a novel feature in RNPs. The two proteins are encoded by different single nuclear genes and no alternatively spliced transcripts could be identified. We constructed a phylogenetic tree for the ten chloroplast CS-RBDs. These results suggest that there is a sizable RNP family in chloroplasts and the diversity was mainly generated through a series of gene duplications rather than through alternative pre-mRNA splicing. The gene for cp29B contains three introns. The first and second introns interrupt the first CS-RBD and the third intron does the second CS-RBD. The position of the first intron site is the same as that in the human hnRNP A1 protein gene.

Nucleic acid-binding specificities of tobacco chloroplast ribonucleoproteins

Many ribonucleoproteins (RNPs) are involved in the regulation of gene expression at the post-transcriptional level. We previously isolated three nuclear-encoded RNPs from tobacco chloroplasts. Here we report their binding specificities for various nucleic acids. The three RNPs synthesized in vitro were subjected to binding assays using ssDNA, dsDNA and four ribonucleotide homopolymers. The affinities for ribonucleotide homopolymers are higher than those for ssDNA and dsDNA, suggesting that they bind preferentially to RNA in vivo. All three RNPs show high specificities for poly(G) and poly(U) in the presence of 1.2-1.8 M NaCl, providing additional evidence for the similarity to HeLa hnRNP proteins. Possible involvement of these RNPs in the post-transcriptional regulation of chloroplast gene expression is discussed.

Nucleic acid binding characteristics of group A/B hnRNP proteins

hnRNP proteins are primarily defined by their specific sedimentational, reconstitutional, and extraction properties and are presumed to be RNA binding. However, it is not clear whether all these proteins have RNA binding capabilities. Recently, using two monoclonal antibodies, fA12 and AC88, we reported that the abundance of a subclass of the highly basic A/B hnRNP proteins was specifically down regulated during terminal differentiation of human and murine cells in vitro. In this report we have examined the nucleic acid binding characteristics of this subclass and other members of the A/B hnRNP proteins in vitro. All members of class A/B hnRNP proteins appear to have similar but not identical nucleic acid binding characteristics. However, the subclass of proteins recognized by AC88 and fA12 exhibit stronger binding affinities and are shown to be highly selective in their binding to RNA vs DNA in vitro. These proteins also preferentially bind poly(U) RNA, suggesting that in vivo they may bind effectively to uridine rich motifs critical in pre-mRNA processing.

A novel heterogeneous nuclear RNP protein with a unique distribution on nascent transcripts

Immediately after the initiation of transcription in eukaryotes, nascent RNA polymerase II transcripts are bound by nuclear proteins resulting in the formation of heterogeneous nuclear ribonucleoprotein (hnRNP) complexes. hnRNP complexes from HeLa cell nuclei contain greater than 20 major proteins in the molecular mass range of 34,000-120,000 D. Among these are the previously described A, B, and C groups of proteins (34,000-43,000 D) and several larger, and as yet uncharacterized, proteins. Here we describe the isolation and characterization of a novel hnRNP protein termed the L protein (64-68 kD by mobility in SDS-polyacrylamide gels). Although L is a bona fide component of hnRNP complexes, it also appears to be a different type of hnRNP protein from those previously characterized. A considerable amount of L is found outside hnRNP complexes, and monoclonal antibodies to the L protein also strongly stain unidentified discrete nonnucleolar structures, in addition to nucleoplasm, in HeLa cell nuclei. Interestingly, the same antibodies stain the majority of nonnucleolar nascent transcripts from the loops of lampbrush chromosomes in the newt, but the most intense staining is localized to the landmark giant loops. The L protein is the first protein of giant loops identified so far, and antibodies to it thus provide a useful tool with which to study these unique RNAs. In addition, isolation and sequencing of cDNA clones for the L protein from human cells predicts a glycine- and proline-rich protein of 60,187 D, which contains two 80 amino acid segments only distantly related to the RNP consensus sequence-type RNA-binding domain. The L protein, therefore, is a new type of hnRNP protein.

Classification and purification of proteins of heterogeneous nuclear ribonucleoprotein particles by RNA-binding specificities

Several proteins of heterogeneous nuclear ribonucleoprotein (hnRNP) particles display very high binding affinities for different ribonucleotide homopolymers. The specificity of some of these proteins at high salt concentrations and in the presence of heparin allows for their rapid one-step purification from HeLa nucleoplasm. We show that the hnRNP C proteins are poly(U)-binding proteins and compare their specificity to that of the previously described cytoplasmic poly(A)-binding protein. These findings provide a useful tool for the classification and purification of hnRNP proteins from various tissues and organisms and indicate that different hnRNP proteins have different RNA-binding specificities.

Classification and purification of proteins of heterogeneous nuclear ribonucleoprotein particles by RNA-binding specificities

Several proteins of heterogeneous nuclear ribonucleoprotein (hnRNP) particles display very high binding affinities for different ribonucleotide homopolymers. The specificity of some of these proteins at high salt concentrations and in the presence of heparin allows for their rapid one-step purification from HeLa nucleoplasm. We show that the hnRNP C proteins are poly(U)-binding proteins and compare their specificity to that of the previously described cytoplasmic poly(A)-binding protein. These findings provide a useful tool for the classification and purification of hnRNP proteins from various tissues and organisms and indicate that different hnRNP proteins have different RNA-binding specificities.

Immunopurification of heterogeneous nuclear ribonucleoprotein particles reveals an assortment of RNA-binding proteins

Heterogeneous nuclear RNA-ribonucleoprotein (hnRNP) particles can be efficiently purified by a specific, rapid, and mild procedure using monoclonal antibodies to hnRNP proteins. We report here on the detailed analysis of the protein composition of immunopurified hnRNP particles from human HeLa cells. By two-dimensional gel electrophoresis, immunopurified hnRNP particles contain at least 24 polypeptides in the range of 34,000-120,000 daltons. The abundant 30,000-40,000 dalton proteins, A, B, and C, described previously, are a subset of these polypeptides. The protein compositions of hnRNP particles found in the nucleoplasm fraction and in the chromatin-nucleolar fraction are very similar. Upon addition of the polyanion heparin, most of the major proteins remain associated in heparin-resistant particles, and only several, mostly minor, proteins dissociate. This provides an aid in the classification of the proteins and an additional criterion for the definition of hnRNP particle components. Chromatography on single-stranded DNA (ssDNA)-agarose in a heparin- and moderate or high salt (higher than 300 mM NaCl)-resistant manner suggests that most, if not all, of these proteins are single-stranded nucleic acid-binding proteins. We describe a general method for the large-scale purification of hnRNP proteins by affinity chromatography on ssDNA columns and its use for the production of new monoclonal antibodies to hnRNP proteins.

Improvements of Western blotting to detect monoclonal antibodies

A comparison of the effects of different factors on the sensitivity of Western blotting technique to detect monoclonal antibodies is described. The major improvements were obtained by: A) renaturating the antigen in the gel before transferring it in carbonate buffer at pH 10 onto nitrocellulose and B) using alkaline-phosphatase-conjugated second antibody instead of peroxidase-conjugated second antibody.