Developmental and tissue-specific expression of mouse pelle-like protein kinase.

The NF-κB/c-Rel proteins are a family of evolutionarily conserved transcription factors activated during development that in the adult, mediate many processes including the immune response. A high degree of sequence similarity is shared between the NF-κB/c-Rel family of transcription factors and the Drosophila Dorsal protein as well as between its cytoplasmic inhibitor, IκBα, and the Drosophila Cactus protein. Genetic analyses of Dorsal have defined components of a signaling pathway for Dorsal activation, including a serine/threonine kinase, Pelle, placed upstream of Dorsal and Cactus. We demonstrate that this pathway is likely to be conserved in mammals by the isolation of a cDNA that encodes a novel mouse protein highly related to Pelle, mPLK (mouse Pelle-like protein kinase). Expression of mPLK mRNA is developmentally regulated in the mouse and in adult tissue mPLK expression is greatest in the liver, a tissue that expresses a high level of NF-κB. Recombinant mPLK produced in bacteria is a protein kinase capable of autophosphorylating and phosphorylating IκBα.

The NF-B/c-Rel proteins are a family of evolutionarily conserved transcription factors activated during development that in the adult, mediate many processes including the immune response. A high degree of sequence similarity is shared between the NF-B/c-Rel family of transcription factors and the Drosophila Dorsal protein as well as between its cytoplasmic inhibitor, IB␣, and the Drosophila Cactus protein. Genetic analyses of Dorsal have defined components of a signaling pathway for Dorsal activation, including a serine/threonine kinase, Pelle, placed upstream of Dorsal and Cactus. We demonstrate that this pathway is likely to be conserved in mammals by the isolation of a cDNA that encodes a novel mouse protein highly related to Pelle, mPLK (mouse Pelle-like protein kinase). Expression of mPLK mRNA is developmentally regulated in the mouse and in adult tissue mPLK expression is greatest in the liver, a tissue that expresses a high level of NF-B. Recombinant mPLK produced in bacteria is a protein kinase capable of autophosphorylating and phosphorylating IB␣.
Acquisition of immune competence in adult mammals depends on a series of developmentally-linked changes in gene expression. How these changes are coordinated during development is unknown. Insight into these processes in mammals can be gained from an understanding of related events in other model developmental systems. The NF-B/c-Rel family of evolutionarily conserved transcription factors are key mediators of genes expressed during activation of immune and inflammatory responses (1)(2)(3). NF-B/c-Rel family members bind DNA as homo-or heterodimers (4 -6), but under basal conditions, NF-B dimers are sequestered in the cytoplasm in a complex with IB␣ (7)(8)(9). In response to cellular stimulation, IB␣ be-comes phosphorylated (1-3,10 -13) and in some instances subjected to proteolytic degradation through a ubiquitin-dependent pathway (14 -16). Removal of IB␣ unmasks a nuclear localization sequence (17)(18)(19), thus allowing NF-B to translocate to the nucleus, bind to its cognate cis-acting element, and transcriptionally activate a variety of genes. IB␣ has been shown to be a substrate for several kinases including protein kinase C, eIF-2 kinase, casein kinase II and a newly described monocyte-specific kinase (1-3, 10 -13). Several of these kinases phosphorylate IB␣ on serine and threonine residues present in the COOH terminus. Phosphorylated serine residues have also been detected in the NH 2 -terminal region of IB␣. Phosphorylation of NH 2 -and COOH-terminal residues has been linked to changes in IB␣ protein stability. Currently, it is unclear whether other kinases are also responsible for IB␣ phosphorylation and whether different NF-B/c-Rel activators operate via distinct kinases.
A high degree of sequence similarity is shared between the NF-B/c-Rel family of transcription factors and the Drosophila Dorsal protein as well as IB␣ and the Drosophila Cactus protein (20). In response to activation of the Toll receptor by Spatzle, a serine/threonine protein kinase encoded by the pelle gene is activated and is believed to phosphorylate Cactus, resulting ultimately in activation of Dorsal. Given the striking parallel between the NF-B/c-Rel and Dorsal pathways, we set out to determine whether a mammalian homolog of the Drosophila Pelle protein existed.

EXPERIMENTAL PROCEDURES
Isolation of mPLK cDNA -Primers, 20 and 21 nucleotides long, were designed to anneal to DNA sequences that encode two regions of the Pelle protein kinase catalytic domain and included degenerate nucleotides of up to 2048-fold degeneracy to ensure efficient priming of unidentified sequences. The sequence of the primer designed to recognize kinase subdomain I: 5Ј-GGIGG(T/A/C/G)TT(C/T)GGIGA(C/T)GT(A/T/C/ G)TA(C/T)-3Ј and the sequence of the primer designed to recognize kinase subdomain VIb: 5Ј-AT(A/G)TTIGCIGG(C/T)TT(A/G/T)AT(A/ G)TC-3Ј. PCR amplifications were performed with mouse embryonic cDNA (1 ng; Clontech Laboratories, Palo Alto, CA) as a template and 100 M of each degenerate primer. After 40 cycles of PCR, products with expected size (400 base pairs) were purified from 1.2% agarose gel using GeneClean II kit (Bio 101, Vista, CA) and cloned using TA cloning kit (Invitrogen, San Diego, CA). cDNA inserts were sequenced with Sequenase according to manufacture's specifications (U. S. Biochemical Corp.). A 400-base pair cDNA fragment was radiolabeled with [␣-32 P]dCTP by primer extension with a multiprime DNA labeling kit (Amersham Corp.). Probes were purified with Quickspin columns (Boehringer Mannheim) and used to screen approximately 1 ϫ 10 6 plaques from a gt11 mouse embryo 5Ј-stretch cDNA library according to the manufacturer's specifications (Clontech Laboratories). After a third round of library screening, 20 positive clones containing inserts ranging from 1500 to 2700 base pairs were isolated. All clones were partially sequenced over the putative catalytic domain, and amino acid sequence comparisons to all available data bases were made with the program BLAST (21). Both strands of the longest clone were sequenced using the Fmol DNA sequencing system (Promega, Madison, WI).
Northern Analysis-Northern blots were purchased from (Clontech) and hybridized according to the manufacturer's specifications with a mPLK cDNA radiolabeled with [␣-32 P]dCTP by primer extension with a Multiprime DNA labeling kit (Amersham). Autoradiograms were prepared by exposing washed membranes to Hyperfilm-MP (Amersham) in the presence of intensifying screens at Ϫ70°C for 2 days.
Cloning, Expression, and Purification of mPLK-The mPLK cDNA was subcloned into the EcoRI site of pET-28a(ϩ) (Novagen, Madison, WI), which generated a His tag at the NH 2 and COOH termini plus an NH 2 -terminal T7 tag to allow for immunodetection. The resulting fu- sion construct was verified by DNA sequencing and used to transform into E. coli strain BL21 (DE-3; Novagen). Induction of mPLK expression and purification of the fusion protein was performed according to the manufacturer's specifications, except that protein production was induced at 30°C overnight with 0.4 mM isopropyl-1-thio-␤-D-galactoside. Cells were harvested by centrifugation, and lysed with a French press in a buffer containing: 5 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl, pH 8.0, 0.1% Triton X-100, 1 g/ml pepstatin A, 1 g/ml leupeptin, 2 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride. Bacterial lysates were centrifuged for 30 min at 39,000 ϫ g, and supernatants were loaded onto Ni 2ϩ -NTA resin (Novagen). His-tagged mPLK was eluted in a buffer containing: 0.5 M NaCl, 20 mM Tris-HCl, pH 8.0, 0.1% Triton X-100, and 200 mM imidazole. Fractions were pooled and concentrated with Centricon-50 filters (Amicon Corp., Beverly, MA). Protein concentrations were determined with the Bio-Rad reagent according to the manufacturer's specifications. The identity of mPLK containing fractions was verified by Western analysis with commercially available T7 tag monoclonal antibody, according to the manufacturer's specifications (Novagen).
Purification of His-tagged IB␣-To generate the His-tagged IB␣ protein E. coli strain BL21 (DE-3), cells containing a fusion gene encoding His-tagged IB␣ were induced at 37°C for 3 h with 0.5 mM isopropyl-1-thio-␤-D-galactoside. Bacteria were collected by centrifugation, resuspended in a buffer containing: 0.5 M NaCl, 50 mM NaP i , 10% glycerol, 0.5% Nonidet P-40, 200 M ␤-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, 2 g/ml pepstatin, sonicated and treated with DNase I (5 mg/ml) for 15 min. Bacterial sonicates were centrifuged (10,000 ϫ g), and the supernatant was applied to a Ni 2ϩ -NTA-agarose column (QIAGEN Inc., Chatsworth, CA). The protein was then eluted with a 30 -300 mM imidazole gradient according to the manufacturer's specifications. Protein determinations were done with a Bio-Rad reagent according to the manufacturer's specifications. The identity of IB␣ containing fractions was verified by Western analysis with commercially available antisera to IB␣ (Santa Cruz Laboratories, Santa Cruz, CA).
In Vitro Kinase Assays-Kinase assays were performed in 20 l of kinase buffer (20 mM Tris-HCl, pH8.0, 50 mM NaCl, 10 mM MgCl 2 , 1 mM dithiothreitol) containing 10 M ATP (10 Ci of [␥-32 P]ATP) and 2 g of purified His-mPLK in the presence or absence of 2 g of purified His-IB␣ at 37°C for 30 min. Individual reactions were terminated by the addition of 2 ϫ SDS-sample buffer (0.1 M Tris, pH 6.8, 0.2 M dithiothreitol, 4% SDS, 0.2% bromphenol blue, 20% glycerol) and boiled 5 min, and proteins were separated on 8% SDS-polyacrylamide gels. Autoradiograms were prepared from dried gels exposed to Hyperfilm-MP (Amersham) film in the presence of intensifying screens at Ϫ70°C for 2 days.

RESULTS AND DISCUSSION
Given the parallel between the NF-B/c-Rel and Dorsal pathways, we set out to determine whether mammals contain a homolog of the Drosophila Pelle protein. Degenerate primers corresponding to the sequences GGFGDVY (kinase subdomain I) and DIKPAN (kinase subdomain VIb) located in the catalytic domain of the Pelle protein kinase (see Fig. 1 legend) were used for PCR amplifications using mouse embryonic cDNA as a template. The DNA sequences of five different PCR products revealed similarity to protein kinases. One sequence encoded an open reading frame that shared 44% amino acid sequence identity with a portion of the Pelle kinase catalytic domain. A mouse embryo cDNA library (Clontech Laboratories, Palo Alto, CA) was then screened using the PCR product as a probe leading to the isolation of a clone containing a 2.7-kilobase insert. This cDNA encodes an open reading frame of 677 amino acids corresponding to a protein with a predicted molecular mass of 74 kilodaltons (Fig. 1, panel A). The first ATG of the open reading frame has an adenine nucleotide (A) in position Ϫ3, a favorable nucleotide and the most common context for initiating translation in vertebrates (22). The 300 amino acids spanning residues 200 -500 of the mouse Pelle-like protein kinase (mPLK) contain the putative protein kinase catalytic domain; the region spanning kinase subdomains I through IX in mPLK is 55% identical to the homologous region in Pelle (20). The COOH-terminal portion of mPLK contains identifia-ble counterparts to the 11 subdomains defined by blocks of identical and similar amino acid residues present in the majority of protein kinases (23). Furthermore, the 15 amino acid residues invariant among all protein kinases are conserved in the mPLK protein sequence (Fig. 1, panel B). There is a high degree of sequence similarity (81%) between mPLK and the recently identified human interleukin-1 receptor-associated kinase (IRAK; Ref. 24). This is consistent with the observation that the cytoplasmic tail of the Drosophila Toll protein and the type 1 interleukin-1 receptor are 43% similar in sequence (25). Two other protein kinases with a high degree of similarity to mPLK are Pti-1 and Pto from Lycopersicon esculentum (26) (Fig. 1). The mPLK amino acid sequence shares 36% identity to Pto and 36% identity to Pti-1 within their kinase domains. The similarity between mPLK, Pto, and Pti-1 is particularly interesting as Pto and Pti-1 belong to a family of plant protein kinase implicated in disease resistance.
There is similarity between the Pelle NH 2 -terminal domain and regions, termed "death" domains, present in the type I tumor necrosis factor receptor and the Fas/ApoI protein re-FIG. 1. Panel A, mPLK amino acid sequence. Panel B, comparison of the amino acid sequences corresponding to mPLK, IRAK, Pelle, Pto, and Pti-1 catalytic domains. The amino acids corresponding to the mPLK catalytic domain is positioned above the corresponding Pelle, Pto, and Pti-1 amino acid sequences (GenBank accession numbers: L08476, U13923, and U28007, respectively). Amino acid numbering is shown to the right; identical amino acids are indicated by dashes. Kinase subdomains are identified by the roman numeral directly above the appropriate region; invariant residues common to all kinases are in boldface type (23). quired for cytotoxic responses (27). Many of the amino acid residues identical between mPLK and Pelle within the NH 2terminal non-catalytic portion of the protein are the same residues that identify the death domain (Fig. 2). Additionally mPLK and the human interleukin-1 receptor-associated kinase have identical death domains with the single exception that mPLK contains a glutamic acid at residue 30, whereas the human interleukin-1 receptor-associated kinase contains a glycine. A putative death domain also exists in the Drosophila Tube protein, which has been shown genetically and biochemically to interact with Pelle (28,29). The death domain has been implicated in protein-protein interactions and would suggest that this region of mPLK may be critical for interacting with another protein of its signaling pathway, such as a Tube homolog.
The pelle gene is expressed throughout the Drosophila life cycle. However, levels of expression are transiently increased in 0 -3-h embryos and are higher in adult females (20). Since the mPLK cDNA was cloned from a day 17 mouse embryo cDNA library, we wanted to determine whether the mPLK gene was expressed at other times during mouse embryogenesis. Analysis of a mRNA isolated from developing mouse embryos revealed no detectable mPLK transcripts in day 7 embryos and a steady increase in the abundance of mPLK transcripts in mRNA isolated from day 11 through day 18 embryos (Fig. 3, left  panel). Northern analysis of mRNA isolated from different adult mouse tissues revealed that, like the pelle gene, the mPLK gene is expressed in adult tissue. Interestingly there is a high level of mPLK gene expression in the liver, a tissue that also expresses a high level of NF-B (30). More moderate levels of mPLK transcripts are present in the kidney and in skeletal muscle (Fig. 3, right panel).
The fact that mPLK resembled a protein kinase led us to determine whether it had protein kinase activity. Results of in vitro kinase assays with bacterially expressed His-tagged mPLK revealed that mPLK is a protein kinase, as judged by its ability to autophosphorylate (Fig. 4, lanes 3 and 4). Pelle has been implicated in the control of the Dorsal/Cactus pathway; therefore, we considered the possibility that IB␣, a Cactus homolog, might be a substrate for mPLK in vitro. Recombinant mPLK was able to phosphorylate IB␣ (Fig. 4). Autophosphorylation was reduced in the presence of IB␣ substrate, as is commonly observed with protein kinases (Fig. 4, compare lanes  3 and 4).
The novel mouse protein kinase mPLK appears to be a mammalian homolog of the Drosophila Pelle protein kinase. A family of plant protein kinases highly related to mPLK include not only Pto, which confers resistance to Pseudomonas syringae in tomatoes (26), and Pti-1, which may act downstream of Pto in a regulatory cascade (26), but also Fen, which confers sensitivity to fenthion in tomatoes (31). The relationship between a kinase(s) potentially associated with a mediator of immune and inflammatory responses in mammals, and plant kinases involved in bacterial resistance, suggests not only a high degree of evolutionary conservation but also a high degree of functional commonality for this growing family of kinases.
We have demonstrated that mPLK is capable of autophosphorylation. Like the mammalian MAP kinases, the plant Pelle-like kinases, Pto and Pti-1, are currently thought to function in a regulatory cascade with Pti-1 serving as a substrate for Pto. It will be interesting to determine if mPLK is a homolog of either Pto or Pti-1 or instead operates at yet another point in the signaling pathway. The nature of the activating signal is of great interest. Based upon the parallels between the Drosophila Dorsal protein and NF-B, one potential mPLK substrate is IB␣. We demonstrate IB␣ phosphorylation in vitro by recombinant mPLK. The functional significance of the IB␣ phosphorylation remains to be determined. However these results coupled with the autophosphorylation of mPLK demonstrate that mPLK can function as a kinase. The role of the putative death domain in the activation of mPLK through its signaling pathway also needs to be evaluated.
The link between mPLK and NF-B/IB␣ is of prime physiological importance. The acute phase response is the body's primary defense against bacterial, viral, and parasitic infection (32). The expression of numerous acute phase response genes, including those encoding interleukin-6 and -8 is modulated by NF-B/c-Rel family members (33,34). Consistent with a role for mPLK in the control of genes encoding acute phase proteins, mPLK gene expression was highest in the liver, the primary site of acute phase protein synthesis in vivo. Intriguingly, massive degeneration of the liver via programmed cell death or apoptosis occurs in mice lacking the NF-B family member, RelA (35). Developmentally, de novo expression of the mPLK gene occurred near the time that hematopoiesis is initiated in the fetal liver (36). The functional significance of this latter finding is unclear; however, in parallel to the RelA/p65 null phenotype, it could reflect the initiation of mPLK expression during the early stages of liver organogenesis. Alternatively, expression by day 11 of development could reflect de novo expression of mPLK in monocyte/macrophages, another site of acute phase protein synthesis. Our data suggests that mPLK is expressed in a tissue-specific manner, yet the NF-B/c-Rel family of transcription factors are ubiquitously expressed. It will be of great interest to determine if mPLK, like the plant kinases Pto, Pti-1, and Fen, is a member of a larger family of kinases in mammals with distinct patterns of tissue distribution. Based upon amino acid homology, the recently identified human interleukin-1 receptor-associated kinase is a highly related family member (24). Therapeutically this class of kinases could be prime targets for the management of acute and chronic inflammatory conditions.