In plant breeding, a great challenge is the collection or development of a large amount of germplasm resources. To this end, several strategies, such as T-DNA or transposon insertional mutagenesis [1] -[3] chemically-induced mutagenesis [4] [5] , have been used to develop different germplasms. However, the mutational spectrum of insertional mutagenesis with effect on gene function is mostly limited to gene knock-out disruptions, which often resulted in failure of functional dissection of mutants when mutations happen to lethal or highly pleiotropic genes [6] . Furthermore, the size of saturated populations is extremely large because each line carries only a rather small number of mutations [7] .

Compared with insertional mutation, chemically-induced mutation has been shown some advantages such as high efficiency because each individual line can bear single point missense and nonsense substitutions in hundreds of genes [8] . Therefore, an allelic series of induced mutations with different effects on gene function can be easily isolated by screening a relatively small population of mutated plants [7] . Among the chemical mutagens, ethyl methanesulfonate (EMS) is considered as an effective one because it can from adducts with nucleotides efficiently, resulting in mispairing among these nucleotides with their complementary bases and thus introducing base changes after replication [5] .

Cucumber is one of the most important horticultural crops and is now grown as either a fresh or processed vegetable in the world. Its fruits are rich sources of high quality protein and vitamins which provide an excellent supplement to the lower quality cereal or root and tuber protein consumed in much of the world [9] . Cucumber fruits vary markedly in size, shape and color, and thus always attract attentions from cucumber breeders. However, germplasm resources are still the big bottleneck for the present cucumber breeding. Here we determined the median lethal dose of EMS for cucumber cultivar “Shannong No. 5” and reported cucumber mutants of fruit, flower, tendril and plant size induced by EMS.

Median lethal dose (LD50) is a critical parameter for chemically induced mutagenesis. This value is determined by both mutagen concentration and treatment time and varies in different species and even different cultivars [10] -[13] . Thus we tried to find LD50 of EMS for “Shannong No. 5” at first. With the increase of treatment time by 1% EMS, the number of survival seeds was decreased (Table 1). When the seeds were treated with 1% EMS for 24 h, LD value was 54%, which is very close to LD50. Therefore, the combination of 1% EMS and 24 h was used for subsequent mutagenesis of cucumber.

After induction by 1% EMS for 24 h, the treated seeds (M1) were planted in the greenhouse. Only 79 out of 541 M1 plants produced M2 seeds. Fertility rate of M1 plants was 14.6%. Low fertility is the common phenomenon observed in mutant plants by physical and chemical mutagenesis possibly due to severe damage of mutagens to plant genetic materials [5] [14] . The mutation frequency is one of the most dependable parameters for evaluating the genetic effects of mutagenic treatments [15] [16] . It was estimated by dividing the number of segregating M1 plant progeny with the determinate character by the total number of progeny. For the material grown in bulk, the mutant frequency was estimated by dividing the total number of mutants confirmed by the total number of M2 plants in the bulk population studied [5] . In this study, sixty lines (M2) from M1 cucumbers were raised for further investigation of phenotypes and calculation of mutation frequency. Eleven out of the sixty lines showed mutations including short-fruit mutants, long-fruit mutants, small-flower mutants, big-flower mutants, opposite-tendril mutants and clustered-leaf mutants compared with wild type plants (Figure 2). The mutation frequency was 18.3%. The main reason for this low mutation frequency is that most of the mutants bearing multi-mutational events may be lethal in the first generation [16] . The short-fruit mutants and clustered-leaf mutants showed 1:3 of segregation ratio in M2 populations (Table 2 and Table 3). This ratio is consistent with classic Mendelian model, indicating that these mutant phenotypes may be controlled by a single recessive gene, respectively. Long-fruit phenotype was inherited stably and no segregation was observed in M3 generation (Figure 3). This evidence suggests that this mutant line may be homozygous at M3 generation.

Note:,. The segregation ratio of short-fruit mutant line is consistent with classic Mendelian model.

Note:,. The segregation ratio of clustered-leaf mutant line is consistent with classic Mendelian model.

All these mutant lines are now still under further investigation and we believe that these mutants can contribute greatly to our future cucumber breeding.

This work was supported by Research Award Fund for Outstanding Middle-aged and Young Scientist of Shandong Province (NO. BS2011NY010), Research Fund for the Doctoral Program of Higher Education of China (20113702120008) and the China Agriculture Research System (CARS-25-D).