To identify potential spa type changes, the spa types from the 319 patients with repeated

MRSA occurrence were analyzed (Fig. 1). Ninety-four percent of the patients had MRSA of only one spa type while 20 patients had two spa types. No one had more than two types. Of these 20 patients, seven had spa types so different from each other that they were considered to be two independent MRSA acquisitions. The remaining 13 patients had two spa types that were closely related. The gender and age, spa type and body site of MRSA isolate, and time between sampling for these isolates is shown in Table 1. The age of the patients varied from 19 to 90. Between two and 17 MRSA isolates were recovered from each patient. The time from the first to the last recording of MRSA was hypoxia-inducible factor pathway between 0 (simultaneously) and 22 months. Out of 88 isolates, the largest group (40) was from skin and soft tissue infection, 23 samples from the nose and 13 from the throat. To further evaluate whether the paired isolates from the same patient were clonally related, a multiple-locus variable number tandem repeat analysis (MLVA) was performed,

based on the method described by Schouls et al. (2009). The discriminatory power of this MLVA is the same as pulsed-field gel electrophoresis and MLVA types can be clustered into MLVA complexes, which coincide with MLST clonal Akt inhibitor complexes (Schouls et al., 2009). In the present study, the MLVA was performed with seven primer pairs (spa-primers were excluded). For PCR, fluorescent dyes were omitted. PCR bands were fragmented on a Qiaxcel

System (Qiagen, Hilden, Germany) and band sizes were determined using the qiaxcel biocalculator software. MLVA analysis was performed on the first isolate of Ancestor and Variant spa types. Thirteen of 319 patients Thalidomide (4%) with a total of 30 MRSA isolates had two spa types with changes that could be assigned to mutational events, suggesting clonal microevolution in the spa repeat region. Twelve different events of spa type change were found; one change was found in two individuals (Table 2). MLVA analysis confirmed that all pairs of isolates were clonally related. All pairs of Ancestor and Variant exhibited identical band patterns except one case with a single band difference (data not shown). The most common mutational event detected was a deletion of spa repeats (10 events), followed by repeat duplication (three events) and point mutation (one event) (Table 2). This was in agreement with the changes found by Kahl et al. (2005) and Sakwinska et al. (2010). Between one and nine repeats were deleted. In two cases (patients 7 and 9), the deletion seemed to be not of a repeat but of 24 bps spanning two repeats, thereby creating a new repeat from parts of the two original repeats. Three changes probably involved repeat duplication. In the first case, spa types t005 and t1276 were involved (patient 3; Table 2). Because t005 is often found in Copenhagen, we consider it to be the Ancestor.