Maize CFM2 Is Localized to the Chloroplast Stroma.
Chloroplasts (Cp) and chloroplast subfractions were from the fractionated chloroplast preparation described previously (Williams and Barkan, 2003). Equivalent proportions of each fraction were examined by immunoblot analysis using antisera specific for the protein indicated to the left. The same blot stained with Ponceau S is shown below. Cpn60, PsbA, and IM35 were used as markers for the stroma, thylakoid, and inner envelope, respectively. The major band detected with the CFM2 antibodies corresponds in size to the CFM2 protein with four CRM domains that is conserved with the gene models of the Arabidopsis and rice orthologs. The asterisk denotes a second protein detected by affinity-purified antibodies from two different rabbits immunized with the CFM2 antigen. Env, envelope; Str, stroma; Thy, thylakoid.

CFM2 Is Found in Large RNPs That Include CAF1 and/or CAF2.
(A) Immunoblots showing coimmunoprecipitation of CFM2 with CAF1 and CAF2. Immunoprecipitations were performed with chloroplast stroma, using the antibody named at the top (IP Ab); antibodies from two different rabbits immunized with the CFM2 antigen were used in replicate assays (lanes 1 and 2). The presence of specific proteins in the immunoprecipitation pellets was determined by probing replicate blots with the antibodies listed to the left. All antibodies were affinity purified against their cognate antigen prior to use. An anti-OE23 immunoprecipitation served as a negative control. The asterisk marks the ∼80-kD protein detected with the CFM2 antisera.
(B) Cosedimentation of CFM2 with CAF1, CAF2, and intron RNAs. Stromal extract was sedimented in sucrose gradients under conditions in which particles greater than ∼60S pellet. An equal proportion of each fraction was analyzed by probing immunoblots with the antibody indicated to the left. The Ponceau S stained blot illustrates the position of Rubisco in the gradient; RbcL is the large subunit of Rubisco. Bottom panels show RNA gel blots of RNA purified from the same gradient fractions. The probes were intron specific and are indicated to the left. The methylene blue–stained blot illustrates the position of ribosomal subunits in the gradient: 16S rRNA marks 30S ribosomal subunits, and 23S* rRNAs mark 50S ribosomal subunits.

Immunoblot Analysis of Subunits of Photosynthetic Complexes in At CFM2 Mutants.
Total proteins from seedling leaves (10 μg or dilutions as indicated) were analyzed on immunoblots by probing with antibodies specific for the proteins indicated to the right. The Ponceau S–stained filter at the bottom illustrates the relative loading of the samples and the level of RbcL, the large subunit of Rubisco. PSI, photosystem I.

Chloroplast Splicing Factors and Their Intron Targets in Land Plants.
The group II introns in land plant chloroplasts are divided into subgroups IIA and IIB according to Michel et al. (1989). Those found in Arabidopsis but not in maize are marked with asterisks. Nucleus-encoded splicing factors, shown to the outside, are annotated with their conserved domains. Solid arrows mark interactions supported both by splicing defects in mutant backgrounds and by RNA coimmunoprecipitation data. Dashed arrows mark interactions supported by either genetic or coimmunoprecipitation data, but not both. Where analyzed (this work; Asakura and Barkan, 2006), the functions of these proteins are conserved between maize and Arabidopsis. A role for the plastid-encoded protein MatK in subgroup IIA splicing has been proposed but has not been confirmed (reviewed in Bonen and Vogel, 2001). HCF152 is an Arabidopsis protein that is required for the accumulation of spliced petB RNA but not for the accumulation of excised intron (Meierhoff et al., 2003). Results are summarized from this work and from Jenkins et al. (1997), Vogel et al. (1999), Till et al. (2001), Ostheimer et al. (2003), Schmitz-Linneweber et al. (2005, 2006), and Watkins et al. (2007).

The Maize cfm2 Gene.
(A) Domain organization of the CFM2 protein. Regions of similarity with CRS1 are represented with like shading. The fragment used for antisera generation is marked.
(B) Schematic of the major maize cfm2 mRNA. Exons are indicated with boxes and introns with dotted lines. Exons in black were represented in cDNAs. The gray exon segment at the 5′ end was inferred from genomic DNA sequence. The exons represented in the cDNA probes used for DNA and RNA gel blots are indicated.
(C) RNA gel blot hybridization showing cfm2 mRNA. Seedling leaf RNA (polyA+, polyA-, or total) was hybridized to cDNA probe B diagrammed in (B). The same blot stained with methylene blue is shown below, with rRNA bands marked.
(D) DNA gel blot hybridization suggesting the presence of two cfm2 genes in maize. Total leaf DNA digested with the indicated enzyme was hybridized with cDNA probes A or C, diagrammed in (B). The blot was washed at high stringency (65°C, 0.2× SSC). The genomic regions corresponding to these probes lack sites for the enzymes used to digest the DNA.

Identification of CFM2's RNA Ligands in Coimmunoprecipitation Assays.
(A) Summary of RIP-chip data. The median log₂-transformed enrichment ratios (F635/F532) for four replicate CFM2 RIP-chip assays (two with each of two CFM2 antisera) are plotted according to chromosomal position (black line). The median ratios for two negative control immunoprecipitations (one with antibody to OE16 and one with antibody to OE23) are shown for comparison (gray line). Fragments for which fewer than two replicate spots per array yielded an F532 signal above background were excluded and appear as gaps in the curve.
(B) Validation of CFM2's RNA ligands by slot blot hybridization. RNA purified from the pellets and supernatants of immunoprecipitations with antisera to the indicated proteins was hybridized to the indicated intron-specific probes. Immunoprecipitations with CRS1 and OE23 antisera and probing for the trnG-UCC intron served as negative controls. The two anti-CFM2 immunoprecipitations used antibodies from different immunized rabbits. The same fraction of the pellet and supernatant samples was applied to the membrane. Differences in the maximal enrichment observed for ycf3-int1, ndhA-int, and trnL-int are not unexpected because the fraction of the RNA pool bound to CFM2 could differ in each case.

At CFM2 Insertion Mutants.
(A) T-DNA insertions in At CFM2 (At3g01370). Exons and introns are indicated with black rectangles and thin lines, respectively. The positions of the T-DNA insertions in At cfm2-1 (SALK_000703) and At cfm2-2 (SAIL_67_E05) are indicated by filled triangles. Arrows show the primers used for RT-PCR (D). LB, T-DNA left border.
(B) Phenotypes conditioned by At CFM2 mutations. Siliques resulting from self-pollination of an At cfm2-1/+ plant are shown at top. White seeds in a young silique and brown shriveled seeds in an older silique are indicated with arrows. These shriveled seeds germinate poorly; those that germinate yield stunted roots and either no shoot (data not shown) or albino, stunted shoots as shown in (C). Shown below are the phenotypes of a homozygous At cfm2-2 mutant and a heteroallelic plant harboring both mutant alleles. Plants were grown for 17 d on MS medium containing 2% sucrose in cycles of 14 h light/10 h dark. Bars = 5 mm.
(C) Phenotype of rare At cfm2-1 homozygote that developed leaves. Plants were grown for 25 d on MS medium containing 2% sucrose in cycles of 10 h light/14 h dark. An At cfm2-2 homozygote grown in parallel is shown for comparison. Bars = 5 mm
(D) RT-PCR analysis of At CFM2 mRNA in At CFM2 mutants. Two hundred nanograms of total leaf RNA from plants with the indicated genotype (homozygous unless otherwise indicated) was analyzed by RT-PCR using the primer pairs indicated to the left and diagrammed in (A). At crs1-1 mutants were analyzed to show the level of At CFM2 mRNA in a different nonphotosynthetic mutant. Dilutions of the wild-type sample were analyzed to illustrate the relationship between signal intensity and input RNA level. RT-PCR of actin 2 mRNA was used as an internal control. The PCR products were detected by staining with ethidium bromide. –RT, control reaction lacking reverse transcriptase.

Chloroplast RNA Splicing Defects in At CFM2 Mutants.
(A) Poisoned-primer extension assays monitoring splicing of chloroplast group II introns. Total leaf RNA (10 μg) from wild-type, At cfm2-2, At cfm2-1/At cfm2-2, and At crs1-1 seedlings was analyzed. The length in nucleotides of each predicted product is indicated. The size of the unlabeled product in the ycf3-int1 assay matches that predicted for the product from an unspliced RNA template in which the last nucleotide of the intron (a cytidine residue) is edited. U, unspliced; S, spliced; P, primer.
(B) RNA gel blots confirming a ycf3-int1 splicing defect in At CFM2 mutants. Total leaf RNA (5 μg) was probed with exon 2 or intron 1 specific probes, as indicated. I, excised intron; S, spliced mRNA; U, unspliced precursor that includes intron 1.
(C) RNA gel blots demonstrating trnL-UAA splicing defect in At CFM2 mutants.
Leaf RNA (1 μg) from seedlings with the indicated genotypes was analyzed by RNA gel blot hybridization using a probe that includes both exon and intron sequences. The same blots stained with methylene blue are shown below; 25S and 18S are cytosolic rRNAs. The ratio of unspliced-to-spliced RNA was quantified with a phosphor imager and is shown in the bar graphs. U, unspliced; S, spliced.
(D) Analysis of the splicing of chloroplast introns found in Arabidopsis but not in maize. Poisoned-primer extension assays were used to monitor clpP-int1 and clpP-int2 splicing, as described in (A). The unlabeled band in the clpP-int2 assays may represent a structure-induced reverse transcriptase stop within the intron, as its abundance is proportional to that of the full-length product from an unspliced template. A ribonuclease protection assay using the diagrammed probe was used to monitor rpoC1 splicing. U, unspliced; S, spliced; P, primer.