Table S1. Primers used for qPCR analysis.

Table S2. Potential genetic changes between 10G, 10G-0.2 and 10G-0.6 as detected by Illumina next generation whole genome sequencing.

Fig. S1. Transfection experiments with plasmids containing BSD resistance markers.

A. Schematic of the 3.2-1371-LH-bsdR and BsdR plasmids. Luc refers to the luciferase ORF and bsd refers to the Blasticidin S deaminase ORF. The 3.2-1371-LH-bsdR plasmid was derived from the 3.2-1371-LH plasmid (Crowley et al., 2011) by inserting a SmaI-BglII fragment of the pHBupsCR plasmid (Voss et al., 2006) containing the BSD deaminase cassette and rep20 repeats into unique SmaI-BamHI sites downstream of the luciferase expression cassette. To generate the BsdR plasmid, the luciferase expression cassette containing the clag3.2 upstream region was removed by digestion with KpnI and religation.

B. Transcript levels of clag3.1 and clag3.2 in unselected control 10G cultures and the same line transfected with the plasmid 3.2-1371-LH-bsdR and maintained under BSD selective pressure for over 2 months. Results were normalized against rhoph2 expression as in Fig. 1B in the main text.

Fig. S2. Expression of clag2, total clag3, clag8 and clag9 in parasites selected with BSD.

A. Expression levels for clag2, total clag3 (measured with primers that recognize a region that is identical between clag3.1 and clag3.2), clag8 and clag9 were determined in the same samples as in Fig. 1B and C in the main text. Note the different scale of the y-axis in the clag2 panel. Results were normalized against rhoph2 expression as in Fig. 1B.

B. Transcript levels of clag2 in 1.2B subclone selected for 2 weeks with 0.3 μg ml−1 BSD (1.2B-0.3), and then 4 weeks with 0.6 μg ml−1 BSD (1.2B-0.6), or grown in parallel in the absence of drug (1.2B). Expression levels in 10G are shown for comparison.

Fig. S3. Changes in clag3 expression are not associated with genetic rearrangements at the clag3 loci or altered copy number of clag3 genes.

A. Long-PCR analysis of the clag3 loci as previously described (Iriko et al., 2008). The combinations of primers in the two upper panels amplify the full unaltered clag3.1 and clag3.2 loci respectively. The primer combinations in the two lower panels would yield a PCR product only if recombination events had occurred between the two loci (Iriko et al., 2008).

B. qPCR analysis of genomic DNA with primers recognizing clag3 genes. Values were normalized against the single copy essential gene seryl tRNA synthetase. Normalization against rhoph2 yielded almost identical results (data not shown).

Fig. S4. Illumina next generation sequencing of 10G (unselected), 10G-0.2 and 10G-0.6 in the genomic region where clag3 genes are located. Snapshot of the reads in the clag3 region aligned to the 3D7 reference genome (v. 9.3), as viewed in the Integrative Genomics Viewer (IGV) (Robinson et al., 2011; Thorvaldsdottir et al., 2013). For each parasite line, the upper panel represents read coverage (scale 0-116, 0-276 and 0-163 for 10G, 10G-0.2 and 10G-0.6 respectively), whereas the lower panel shows the position of individual reads. Salmon and blue colours indicate reads corresponding to the two different DNA strands, whereas grey indicate non-unique reads. Non-unique reads corresponding to sequences that are identical between clag3.1 and clag3.2 were split between the two genes (a mutation occurring within non-unique regions in only one of the two genes would be expected to reach a maximum frequency of 50%). SNPs that reach a frequency above 20% are indicated in the coverage panels. The single mutation identified within the clag3 region, which occurred in 10G-0.6, is indicated by an arrowhead. This mutation also occurred in control 10G and 10G-0.2, but it is not indicated in their coverage panels because the frequency of the alternative allele was slightly below the 20% threshold (19% in 10G and 18% in 10G-0.2, compared with 38% in 10G-0.6). Furthermore, this SNP is located next to a string of A's. Hence, this mutation is unlikely to be associated with BSD selection.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.