Phenotypic Mutation 'bluebonnet' (pdf version)
Allelebluebonnet
Mutation Type missense
Chromosome2
Coordinate164,904,221 bp (GRCm39)
Base Change T ⇒ C (forward strand)
Gene Cd40
Gene Name CD40 antigen
Synonym(s) Tnfrsf5, p50, Bp50, Cd40
Chromosomal Location 164,897,535-164,913,574 bp (+) (GRCm39)
MGI Phenotype FUNCTION: [Summary is not available for the mouse gene. This summary is for the human ortholog.] This gene is a member of the TNF-receptor superfamily. The encoded protein is a receptor on antigen-presenting cells of the immune system and is essential for mediating a broad variety of immune and inflammatory responses including T cell-dependent immunoglobulin class switching, memory B cell development, and germinal center formation. AT-hook transcription factor AKNA is reported to coordinately regulate the expression of this receptor and its ligand, which may be important for homotypic cell interactions. Adaptor protein TNFR2 interacts with this receptor and serves as a mediator of the signal transduction. The interaction of this receptor and its ligand is found to be necessary for amyloid-beta-induced microglial activation, and thus is thought to be an early event in Alzheimer disease pathogenesis. Mutations affecting this gene are the cause of autosomal recessive hyper-IgM immunodeficiency type 3 (HIGM3). Multiple alternatively spliced transcript variants of this gene encoding distinct isoforms have been reported. [provided by RefSeq, Nov 2014]
PHENOTYPE: Homozygous inactivation of this gene may cause impaired immunoglobulin class switching and germinal center formation, reduced susceptibility to type II hypersensitivity reaction, impaired priming of T cells and control of M. tuberculosis infection, and altered response to transplant. [provided by MGI curators]
Accession Number

NCBI RefSeq: NM_011611, NM_170703, NM_170704, NM_170702; MGI:88336

MappedYes 
Amino Acid Change Cysteine changed to Arginine
Institutional SourceBeutler Lab
Gene Model predicted gene model for protein(s): [ENSMUSP00000017799] [ENSMUSP00000073386] [ENSMUSP00000080059] [ENSMUSP00000122981] [ENSMUSP00000138707] [ENSMUSP00000139193]
AlphaFold P27512
SMART Domains Protein: ENSMUSP00000017799
Gene: ENSMUSG00000017652
AA Change: C23R

DomainStartEndE-ValueType
signal peptide 1 23 N/A INTRINSIC
TNFR 26 59 9.45e-6 SMART
TNFR 62 103 2.38e-11 SMART
TNFR 105 143 4.55e-8 SMART
TNFR 146 186 2.42e-3 SMART
transmembrane domain 193 215 N/A INTRINSIC
Predicted Effect probably benign

PolyPhen 2 Score 0.013 (Sensitivity: 0.96; Specificity: 0.78)
(Using ENSMUST00000017799)
SMART Domains Protein: ENSMUSP00000073386
Gene: ENSMUSG00000017652
AA Change: C23R

DomainStartEndE-ValueType
signal peptide 1 23 N/A INTRINSIC
TNFR 26 59 9.45e-6 SMART
TNFR 62 103 2.38e-11 SMART
TNFR 105 143 4.55e-8 SMART
TNFR 146 186 2.42e-3 SMART
Predicted Effect probably benign

PolyPhen 2 Score 0.013 (Sensitivity: 0.96; Specificity: 0.78)
(Using ENSMUST00000073707)
SMART Domains Protein: ENSMUSP00000080059
Gene: ENSMUSG00000017652
AA Change: C23R

DomainStartEndE-ValueType
signal peptide 1 23 N/A INTRINSIC
TNFR 26 59 9.45e-6 SMART
TNFR 62 103 2.38e-11 SMART
TNFR 105 143 4.55e-8 SMART
Predicted Effect probably benign

PolyPhen 2 Score 0.013 (Sensitivity: 0.96; Specificity: 0.78)
(Using ENSMUST00000081310)
SMART Domains Protein: ENSMUSP00000122981
Gene: ENSMUSG00000017652
AA Change: C61R

DomainStartEndE-ValueType
signal peptide 1 19 N/A INTRINSIC
TNFR 64 97 9.45e-6 SMART
Blast:TNFR 100 121 1e-8 BLAST
Predicted Effect probably benign

PolyPhen 2 Score 0.226 (Sensitivity: 0.91; Specificity: 0.88)
(Using ENSMUST00000140951)
Predicted Effect probably benign
SMART Domains Protein: ENSMUSP00000139193
Gene: ENSMUSG00000017652
AA Change: C23R

DomainStartEndE-ValueType
signal peptide 1 23 N/A INTRINSIC
TNFR 26 59 9.45e-6 SMART
TNFR 62 103 2.38e-11 SMART
TNFR 105 143 4.55e-8 SMART
TNFR 146 186 2.42e-3 SMART
Predicted Effect probably benign

PolyPhen 2 Score 0.013 (Sensitivity: 0.96; Specificity: 0.78)
(Using ENSMUST00000184221)
Meta Mutation Damage Score 0.0898 question?
Is this an essential gene? Non Essential (E-score: 0.000) question?
Phenotypic Category Autosomal Recessive
Candidate Explorer Status loading ...
Single pedigree
Linkage Analysis Data
Penetrance  
Alleles Listed at MGI

All mutations/alleles(8) : Gene trapped(1) Targeted(7)

Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
noelle UTSW 2 164905483 critical splice donor site probably null
R0553:Cd40 UTSW 2 164912661 missense probably benign 0.01
R1115:Cd40 UTSW 2 164912681 missense possibly damaging 0.59
R1134:Cd40 UTSW 2 164912738 missense probably benign 0.44
R2036:Cd40 UTSW 2 164904221 missense probably benign 0.23
R2938:Cd40 UTSW 2 164911622 missense probably benign 0.01
R3034:Cd40 UTSW 2 164904235 missense probably benign 0.02
R4690:Cd40 UTSW 2 164911615 missense possibly damaging 0.68
R5222:Cd40 UTSW 2 164908464 missense probably benign 0.41
R5310:Cd40 UTSW 2 164905483 critical splice donor site probably null
R7318:Cd40 UTSW 2 164904255 missense possibly damaging 0.51
R7833:Cd40 UTSW 2 164908431 missense probably benign 0.01
R7905:Cd40 UTSW 2 164904245 missense probably damaging 1.00
R8069:Cd40 UTSW 2 164898695 missense unknown
R8371:Cd40 UTSW 2 164908458 missense probably damaging 1.00
R9177:Cd40 UTSW 2 164905465 missense probably damaging 1.00
R9224:Cd40 UTSW 2 164898716 missense unknown
R9311:Cd40 UTSW 2 164912667 missense possibly damaging 0.87
R9419:Cd40 UTSW 2 164904162 intron probably benign
R9625:Cd40 UTSW 2 164905061 missense probably benign
Z1088:Cd40 UTSW 2 164904960 missense probably damaging 0.96
Mode of Inheritance Autosomal Recessive
Local Stock
MMRRC Submission 038194-MU
Last Updated 2019-09-04 9:45 PM by Katherine Timer
Record Created 2015-04-09 7:27 PM by Jin Huk Choi
Record Posted 2015-05-14
Phenotypic Description

Figure 1. Homozygous bluebonnet mice exhibit diminished T-dependent antibody responses to to ovalbumin administered with aluminum hydroxide. IgG levels were determined by ELISA. Normalized data are shown. Abbreviations: WT, wild-type; REF, homozygous reference mice; HET, heterozygous variant mice; VAR, homozygous variant mice. Mean (μ) and standard deviation (σ) are indicated.

The bluebonnet phenotype was identified among N-ethyl-N-nitrosourea (ENU)-mutagenized G3 mice of the pedigree R2036, some of which exhibited a null T-dependent antibody response to ovalbumin administered with aluminum hydroxide (Figure 1).

Nature of Mutation

Figure 2. Linkage mapping of reduced T-dependent antibody responses to to ovalbumin administered with aluminum hydroxide using a recessive model of inheritance. Manhattan plot shows -log10 P values (Y-axis) plotted against the chromosome positions of 63 mutations (X-axis) identified in the G1 male of pedigree R2036. Normalized phenotype data are shown for single locus linkage analysis with consideration of G2 dam identity.  Horizontal pink and red lines represent thresholds of P = 0.05, and the threshold for P = 0.05 after applying Bonferroni correction, respectively.

Whole exome HiSeq sequencing of the G1 grandsire identified 63 mutations. The reduced T-dependent antibody response phenotype was linked by continuous variable mapping to a mutation in Cd40:  a T to C transition at base pair 165,062,301 (v38) on chromosome 2, or base pair 6,698 in the GenBank genomic region NC_000068 encoding Cd40. Linkage was found with a recessive model of inheritance (P = 1.178 x 10-6), wherein 2 variant homozygotes departed phenotypically from 8 homozygous reference mice and 6 heterozygous mice (Figure 2). The mutation corresponds to residue 139 in the mRNA sequences NM_011611 (variant 1; within exon 2 of 9 total exons), NM_170703 (variant 2; within exon 2 of 8 total exons), NM_170704 (variant 4; within exon 2 of 9 total exons), and NM_170702 (variant 5; within exon 2 of 8 total exons).

6683 GTCCATCTAGGGCAGTGTGTTACGTGCAGTGAC

18   -V--H--L--G--Q--C--V--T--C--S--D-

Genomic numbering corresponds to NC_000068. The mutated nucleotide is indicated in red.  The mutation results in an cysteine (C) to arginine (R) substitution at position 23 (C23R) in all of the isoforms of the CD40 protein, and is strongly predicted by PolyPhen-2 to be benign (score = 0.013) (1).

Illustration of Mutations in
Gene & Protein
Protein Prediction

Figure 3. Domain organization of CD40. The effect of the bluebonnet mutation is indicated. Abbreviations: SP, signal peptide; CRD, cysteine-rich domain; TM, transmembrane domain; cyt-N, cytoplasmic N-terminal; cyt-C, cytoplasmic C-terminal.

CD40 (alternatively, TNFRSF5) is a member of the tumor necrosis factor receptor (TNFR) family that also includes Fas (alternatively, TNFR6; see the record for cherry) and lymphotoxin β receptor (LTβR; see the record for kama). Similar to other members of the TNFR family, CD40 is a single-pass transmembrane-spanning protein with an extracellular domain (ECD; amino acids 24-193; SMART), a transmembrane domain (TMD; amino acids 193-215), and an intracellular domain (ICD; amino acids 216-289) [Figure 3; (2)]. Amino acids 1-23 of CD40 comprise a signal peptide. CD40 has two conserved putative Asn-linked glycosylation sites at Asn153 and Asn180 (2). Within the ECD, CD40 has four cysteine-rich domains [CRDs; amino acids 26-59 (CRD1), 62-103 (CRD2), 105-143 (CRD3), and 146-186 (CRD4)]. CRDs each have six cysteine residues that associate to form three disulfide bonds. CRD1 mediates self-assembly of CD40 and is also required for CD40 expression (3). Dimerization of CD40 is proposed to be required for efficient activation of CD40 and subsequent downstream signaling. CRD2 and CRD3 are required in binding of CD40 ligand (CD40L; see the record for walla). The ICD of CD40 has several motifs that mediate interactions with downstream effectors of several signaling pathways. Amino acids 222-229 are essential for association with Janus kinase 3 (Jak3; see the record for mount tai) (4). Two distinct domains in the cytoplasmic tail, cyt-N (membrane-proximal; (Q/N/R)X(P/A)XEXX(F/W/Y); amino acids 231-238) and cyt-C [membrane-distal; (P/H)(V/I)QE(T/S); amino acids 250-254] bind to different members of the downstream signaling molecule TNF receptor-associated factor (TRAF) family and can activate NF-κB independently (5-7), although both binding sites are required for optimal NF-κB activation (4;8). TRAF1, TRAF2, TRAF3 (see the record for hulk), and TRAF5 bind to cyt-C, whereas TRAF6 binds to cyt-N (4;8;9). Upon CD40 engagement, TRAF2 and TRAF3 directly bind to the cytoplasmic tail of CD40; TRAF1 and TRAF5 are recruited to CD40 through associations with CD40-bound TRAFs (9;10). The crystal structure of a portion of the human CD40 cytoplasmic domain bound to TRAF3 has been solved [PDB:1FLL; (11)]. When bound to TRAF3, the CD40 fragment is a hairpin with two extended segments and a reverse turn centered at amino acids 258-260 (11). P250, Q252, T254 reside within a crevice across the face of the β-sandwich to directly contact TRAF3.  

The bluebonnet mutation results in a cysteine to arginine substitution at amino acid 23 (C23R) within the SP, which may lead to defects in protein secretion. Expression and localization of CD40bluebonnet has not been examined.

Expression/Localization

CD40 is constitutively expressed on the surface of mature B cells, most mature B cell malignancies, some early B cell acute lymphocytic leukemias, monocytes, dendritic cells (DCs), endothelial cells, and epithelial cells (2;4).

Background

Figure 4. Canonical and non-canonical NF-kB signaling pathways. In the canonical pathway, several membrane receptors, including TNFR (tumor-necrosis factor receptor), IL-1R (interleukin-1 receptor) and TLRs (toll-like receptors), signal through kinases and adaptors (TRAFs), resulting in IKK activation. This activation occurs after the K63 ubiquitination of TRAFs and RIP. TAK1 and its adaptor proteins TAB1 and TAB2 bind ubiquitin chains to TRAF and NEMO (IKKγ) resulting in the activation of the IKK complex (NEMO, IKKα and IKKβ). Ubiquitination can be inhibited by deubiquitinating enzymes (DUB). Stimulation of the T-cell receptor (TCR) and B-cell receptor (BCR) results in the recruitment of Src and Syk family kinases. These kinases activate a phosphorylation cascade which leads to the activation of protein kinase C (PKC). The phosphorylation of CARMA1 (CARD (caspase-recruitment domain)-MAGUK (membrane-associated guanylate kinase) protein 1) recruits BCL 10 (B-cell lymphoma 10) and MALT1 (mucosa-associated lymphoid tissue lymphoma translocation gene 1), forming the CBM complex and activating the IKK complex. The IKK complex phosphorylates both IκB, p105 and TPL2 (or MAP3K8), resulting in IκB and p105 ubiquitination and degradation (small pink circles) by 26S proteasome. Degradation of IκB releases activated NF-κB dimers for translocation to the nucleus. A subset of TNFRs such as the lymphotoxin-β receptor (LT-βR),CD40, B-cell-activating factor receptor (BAFFR) and receptor activator of Nf-κB (RANK) can activate the canonical or non-canonical NF-κB signaling pathways. In the non-canonical pathway, the receptors bind to TRAFs to regulate NIK activity. TRAF3 and TRAF2 are recruited to the receptor along with cIAP1/2. TRAF2 undergoes K63 self-ubiquitination and is responsible for the K63 ubiquitination of cIAP1/2. TRAF3 is degraded by K48 ubiquitination, enhanced by the K63 ubiquitination of TRAF2 and cIAP1/2. (Gray arrows represent ubiquitination dependence.) As TRAF levels decrease, NIK is released and phosphorylates IKKα which phosphorylates p100. Phosphorylation and ubiquitination of p100 leads to the 26S proteasomal degradation of p100 and the processing of p52. P52 and RelB are released for translocation to the nucleus. This image is interactive. Click on the image to view mutations in the pathway (red) and the genes affected by these mutations (black). Click on the mutations for more specific information.

The members of the TNFR superfamily regulate cell survival and death through the canonical and non-canonical NF-κB [see the records for Finlay (Nfkb1) and xander (Nfkb2), respectively] signaling pathways. CD40/CD40L provides costimulatory, activating signals to antigen presenting cells (APCs) such as B cells, macrophages, and DCs. The CD40 receptor, expressed on B cells and other APCs, is engaged by CD40L that is present primarily on activated T cells. The canonical and non-canonical NF-κB signaling pathways are activated downstream of TRAFs upon CD40L:CD40 engagement, and are major pathways for transducing the effects of CD40 activation (Figure 4). Activation of NF-κB through the classical pathway depends on degradation of IκB, thus releasing NF-κB to enter the nucleus to coordinate expression of multiple inflammatory and innate immune genes. Stimulation of CD40 also activates the non-canonical NF-κB pathway, which requires IKK-α, but not the IKK complex, and the proteasome-mediated processing of NF-κB2 (p100) to p52 (12;13). CD40 engagement results in recruitment and subsequent degradation of TRAF2 and TRAF3, destabilizing the complex with NF-κB inducing kinase (NIK) (see lucky) and permitting NIK accumulation (14-16). NIK binds and phosphorylates both IKK-α and p100 (17-19). Phosphorylation of p100 on two C-terminal sites triggers its polyubiquitination and proteasomal degradation to p52. Since p52 is typically associated with RelB, activation of the non-canonical NF-κB pathway results in nuclear translocation of p52-RelB dimers (20).

In addition to NF-κB activation, CD40 also mediates its cellular effects by initiating signaling through pathways controlled by p38 MAP kinase (MAPK), c-Jun, STAT (see the record domino for information about Stat1), and Jak3, as well as PI3K and PLC-γ (see the record for queen) [reviewed in (21)]. Jak3 activation results in STAT3 phosphorylation, activation, and translocation to the nucleus. Jak3 is not required for CD40-mediated B cell proliferation, isotype switching, or upregulation of ICAM-1 or CD80 (22). However, pharmacological inhibition of Jak3 prevented the expression of costimulatory molecules and the production of IL-12 in response to CD40 stimulation of DCs (23).

Several functions of CD40:CD40L are described, in brief, below. For an in-depth description of CD40:CD40L functions in humoral and cell-mediated immunity as well as CD40:CD40L-associated signaling, please see the record for walla.

CD40 in humoral immunity

CD40 signaling is essential for the maturation and survival of DCs, resulting in enhanced survival, secretion of cytokines such as interleukin (IL)-1, IL-6, IL-8, IL-10, IL-12, and TNFα, and upregulation of costimulatory molecules including ICAM-1 (24), LFA-3, CD80, and CD86.  Conversely, crosslinking CD40L on T helper (Th) cells contributes to the generation of helper function (25). Once activated, Th cells migrate to the border between the B cell follicles and the T cell zone where they encounter antigen-activated B cells (26).

CD40 induces B cell proliferation (27-29), cell-cell adhesion (30), differentiation (31;32), and tyrosine phosphorylation (33;34); B cells receive a critical costimulatory signal from T cells through CD40. The B cell response to T-dependent antigen involves engagement of CD40 on B cells and subsequent induction of costimulatory molecules (35-37). The concerted action of CD40L and Fas ligand (FasL; see the record for riogrande) on activated T cells promotes clonal proliferation of B cells stimulated by foreign antigen, whereas when B cell receptors have been desensitized by chronic stimulation with self-antigens or have not been bound by antigen CD40L and FasL trigger deletion (38). Lack of CD40L:CD40 signaling completely abolishes T-dependent humoral immunity. Study of mice deficient in either CD40 or CD40L demonstrated that CD40L:CD40 interactions are required for germinal center (GC) formation, progression, and maintenance, as well as antibody isotype switching and affinity maturation, with these processes essential for the generation of memory B cells and long-lived plasma cells (39-42). IgM responses to some T-dependent antigens were found to be impaired in CD40L-deficient mice, suggesting that CD40L:CD40 signaling is not only required for secondary Ig responses but also for primary IgM responses. Interestingly, while no GC response was observed in mice lacking CD40L:CD40 signaling, excessive CD40 stimulation during a primary immune response, either using agonistic CD40 antibodies or elevated T cell help, resulted in ablation of GC formation and premature termination of the humoral immune response, blocked the generation of B cell memory and long-lived bone marrow plasma cells, and diverted all B cells to the plasmablast lineage (43). Thus, a normal GC response requires CD40 signaling to fall within a defined range.

TFH cells also depend on CD40L and IL-21 for their generation and function (44;45). CD40L:CD40 interactions are required for the initiation but not maintenance of B cell differentiation leading to Ig production (46). However, long-lived bone marrow plasma cells decreased their average affinity for antigen subsequent to blockade of CD40L (47), presumably because of disrupted interactions between cognate GC Th cells and B cells during affinity maturation in the GC.

Mutations in CD40 are linked to the autosomal recessive form of immunodeficiency with hyper-IgM [HIGM3; OMIM: #606843); (48-50)].  HIGM syndromes, including HIGM1 (caused by mutations in CD40LG) and HIGM2 (caused by mutations in AID) are characterized by normal to elevated levels of IgM and low levels of IgA, IgG, and IgE, the absence of GCs, and the inability to mount a T-dependent humoral response [reviewed in (51)]. These phenotypes arise from deficient CD40L:CD40 interaction that is required for T-dependent B cell activation, as described above. As a result, HIGM1 patients suffer from recurrent bacterial infections that typically begin between the first and second years of life. These infections include a high incidence of upper and lower respiratory tract infections, otitis media, and pneumonia (52). Patients are also susceptible to developing autoimmunity, neutropenia, and lymphoproliferative diseases, and to opportunistic infections such as cryptosporidium, Pnuemocystis carinii, mycobacteria, and cytomegalovirus (CMV). T cells from a majority of HIGM1 patients exhibit reduced or absent CD40L protein expression and defective Th function despite normal CD40L mRNA levels (53). A high proportion of mutations occur in the extracellular domain (54), resulting in a range of defective binding ability for several CD40L antibodies (53). Both CD40L- and CD40-deficient mice (39-42) and some 50% of HIGM1 patients (52) display normal levels of circulating IgM, suggesting that an increased IgM level is a secondary phenotype and not a genetically determined feature of the disease. Both humans (52;55) and mice (56-58) deficient in CD40 signaling display increased susceptibility to the development of lymphomas and gastrointestinal cancers.  

CD40 in cell-mediated immunity

CD40L and CD40 play an important role in the priming and expansion of antigen-specific Th cells. Adoptively transferred antigen-specific CD4+ T cells lacking CD40L failed to expand upon antigen challenge of the recipients (59). Expression of CD40L on T cells is required for priming of CD4+ T cells. CD40 signaling is not only required for the upregulation of costimulatory molecules and cytokine production by DCs, but also for increasing the expression and stability of the MHC:alloantigen complex (60). Signals from CD40 also promote the survival of DCs by upregulating the anti-apoptotic factor Bcl-XL in an NF-κB-dependent manner (see section on CD40 signaling below) (61)

A fundamental role for CD40L:CD40 signaling in monocyte activation by T cells has also been demonstrated. Engagement of CD40 on macrophages using recombinant CD40L or CD40L-expressing cells resulted in activation of effector function, including production of the proinflammatory cytokines IL-12, IL-6, and TNF-α (62;63). However, CD40L-deficient activated Th cells failed to induce IL-12 or nitric oxide production by macrophages (64-66).

Putative Mechanism

Cd40-deficient (Cd40-/-) mice were deficient in mounting an antigen-specific antibody response or to develop GCs after immunization with T-dependent antigen keyhole limpet hemocyanin (KLH) (67). The T-independent antibody response to 2,4,6-trinitrophenyl-conjugated lipopolysaccharide (TNP-LPS) and TNP-Ficoll in the Cd40-/- mice were normal (67). The loss in T-dependent antibody responses observed in the bluebonnet mice indicates a loss of CD40bluebonnet function including a failure to develop GCs in response to T-dependent antigen immunization.

Primers PCR Primer
bluebonnet_pcr_F: AGTCCTGTGCATCTGTTCGG
bluebonnet_pcr_R: CTTGGGGTATTCTGGCATCAAG

Sequencing Primer
bluebonnet_seq_F: GCTTTGGTAGATGGCAGTAAGAC
bluebonnet_seq_R: TATTCTGGCATCAAGGAAGAGG
Genotyping

PCR program

1) 94°C 2:00
2) 94°C 0:30
3) 55°C 0:30
4) 72°C 1:00
5) repeat steps (2-4) 40x
6) 72°C 10:00
7) 4°C hold


The following sequence of 436 nucleotides is amplified (chromosome 2, + strand):


1   agtcctgtgc atctgttcgg attagagggt tctgcgttct tgctttggta gatggcagta
61  agacgatgtg acaacagagt aaaaaaaaaa atagacctca cactctgggg gctcactttt
121 ctgctttgga tttccacatc agctacagcc tgcgtcttgg ctaactttca acatgccggt
181 ggaagatccc ttccagctgt ccacttctgt ttttaggtcc atctagggca gtgtgttacg
241 tgcagtgaca aacagtacct ccacgatggc cagtgctgtg atttgtgcca gccaggtgag
301 atgctagccc tcctgccccg taccaagacc ctttccctct tggattgctg gtggatgcag
361 accccatata tagactgtga actcaagtct aaaatgaccc atttctccct cttccttgat
421 gccagaatac cccaag


Primer binding sites are underlined and the sequencing primers are highlighted; the mutated nucleotide is shown in red.

References
Science Writers Anne Murray
Illustrators Peter Jurek
AuthorsKuan-Wen Wang, Jin Huk Choi, Bruce Beutler