The coverage of all primers was assessed using Silva (n.d.). Silva provides regularly updated datasets of aligned small (16S/18S, SSU) and large subunit (23S/28S, LSU) ribosomal RNA (rRNA) sequences for all three domains of life (Klindworth et al., 2013; Quast et al., 2013). The latest update, Silva SSU 138.1, was released in December 2019, increasing the number of available SSU sequences to 9,469,124 (Silva, n.d.).

Silva’s TestPrime page¹ was accessed, the primer sequence was entered, the SSU R138.1 Database was selected, the maximum number of mismatches was set to 0, and “Run TestPrime” was selected to allow the system to automatically calculate the primer coverage.

The overall idea of the script is to identify the bases in the primers that do not match the SSU rRNA gene and replace them with degenerate bases. Specifically:

Firstly, the target gene is converted into its reverse complementary sequence, i.e., converting the sense strand to the antisense strand (this step is for the improvement of the reverse primer, while the improvement of the forward primer skips this step). The reverse complementary sequence is achieved through a two-step process: (1) Complementarity: Base pairs are replaced with their corresponding counterparts. A is replaced with T, C is replaced with G. (2) Reverse: The sequence from 5′ to 3′ is rearranged from 3′ to 5′, then treated as a new 5′ to 3′ sequence. For example, the sequence 5’ AGGTAC 3′ has a complementary sequence of 5′ TCCATG 3′, and the reverse complementary sequence is 5’ GTACCT 3′.

Next, locate the corresponding sequences in the gene by searching for bases that match the primer, and identify any mismatched bases. The determination of sequence identity includes exact match, degenerate match, and mismatch. Exact match refers to bases with the same name, such as A matches with A, R matches with R, and so on. Degenerate match refers to degenerate bases matching with the included bases, such as G matches with degenerate bases containing G (e.g., R, K, S, B, V, D, N), and so on. Any cases other than exact match and degenerate match are considered as mismatches. When the number of mismatched bases exceeds 5, it is considered invalid. The threshold for mismatched bases is set at <5 for the following reasons: if there are more than 5 differences, the sequence may not be the primer corresponding sequence; even if the sequence is the corresponding sequence, it means no improvement. This is because the minimum product of degeneracy for five bases is 32, adding one more degenerate base exceeds 64, meaning it would generate 64 different primer sequences, while SILVA recommends that degenerate primers correspond to no more than 60 sequences (Quast et al., 2013).

Finally, bases in the primer that are different from the gene sequence are replaced with degenerate bases, preferably with 2-base degenerate bases, if not possible, 3-base degenerate bases are chosen, and if still not possible, 4-base degenerate bases are chosen. The replacement strategy is as follows: A/G = R (meaning if the different bases between the primer and the gene are A and G, then the primer bases are changed to degenerate base R), C/T = Y, A/C = M, G/T = K, G/C=S, A/T = W, A/Y=H, T/M = H, C/W=H, G/Y=B, T/S=B, C/K=B, G/M = V, A/S=V, C/R = V, G/W=D, A/K=D, T/R = D, A/B=N, T/V=N, C/D=N, G/H=N.

The steps for using the script in conjunction with the Silva database are as follows: Step 1: Prepare the SSU rRNA gene. Evaluate the universal primers in the Silva database, and then download a SSU rRNA gene sequence from the target microorganism that is not covered by the universal primers and place it in the “gene” folder. The primer coverage assessment and gene download process are illustrated in Figure 1. Step 2: Prepare the primers. Place the forward primer (F primer) and reverse primer (R primer) of the universal primers to be improved into the “old F” and “old R” folders, respectively. Step 3: Run the commands. Execute the “script F” or “script R” command in the macOS terminal or Linux shell. Step 4: Collect the new primers. The new primers will be displayed in the running interface and in the “new F/R” folder. Replace the old primers with the new ones and repeat steps 1–4 for iterative improvement until Silva indicates that the new primers are invalid due to containing multiple degenerate bases, resulting in more than 60 primer sequences. The last round of effective primers is considered the final improved primers. The script usage process is illustrated in Figure 2.

This study selected 8 pairs of classic primers as examples to demonstrate the personalized improvement using the “Degenerate primer” tool to enhance the coverage of target microorganisms. Microorganisms with coverage below 100% were considered target microorganisms, and this study only selected some of microorganisms with low coverage. For convenience, this study renamed these primers using B, A, and E to represent bacteria, archaea, and eukaryotes, respectively, with names indicating the start position of the F primer and the end position of the R primer on the forward strand of the SSU rRNA gene. These primers and their corresponding microorganisms are as follows:

Using both the original primers and the improved primers to amplify the microbial SSU rRNA genes of the same sample can be employed to compare the impact of primer coverage on actual detection results. To ensure the presence of the targeted microorganisms in the samples, this study extracted DNA from 8 typical environmental samples (soil, sediment, soil-derived cultures, and sediment-derived cultures) known to contain Dehalococcoides. The improved primers specific for Dehalococcoides BA-515F-806R-M1 and the original primers BA-515F-806R were used to amplify and sequence the same SSU rRNA gene from the same sample for comparison.

The microbial DNA in the environmental sample test groups was extracted using a PowerSoil^® DNA Isolation Kit (Mobio Laboratories, Inc., Carlsbad, CA, United States) according to the manufacturer’s instructions. The SSU rRNA gene was amplified using the two pairs of primer and sequenced by the Illumina MiSeq platform. The Thermal cycling program is set as follows: 95°C for 3 min; 95°C for 30 s, 55°C for 30 s, 72°C for 45 s, a total of 25 cycles; 72°C for 5 min; Hold at 4°C. The sequencing results were processed by the Shanghai Shenggong Biotechnology Co., Ltd. and related software Usearch et al. to obtain the final OTU information (Edgar, 2013, 2016), which was then prepared for the downstream data analysis.

This study compared the differences in the number of detected microbial species between the unimproved primer BA-515F-806R and the improved primer BA-515F-806R-M1 when amplifying the same sample using OTU counts. Additionally, the study investigated the differences in the number of detected species at the genus level for g_ Dehalococcoides and nine other genera (g_Inhella, g__Methylomicrobium, g_Caminicella, g_Myroides, g_Dokdonella, g_Desulfovibrio, g_Pedomicrobium, g_Lewinella, g__Turicibacter), which exhibited increased coverage with BA-515F-806R-M1, between amplification with the improved and unimproved primers in the same sample.

Raw data of high throughput sequencing of SSU rRNA gene have been uploaded to the NCBI Sequence Read Archive database under BioProject ID PRJNA1047931.

This study developed a script to improve universal primers, named “Degenerate primer 111.” This tool, used in conjunction with the Silva website, aligns “1” universal primer to “1” target microbial SSU rRNA gene not covered by the universal primer, generating “1” new universal primer covering the target gene. Iterative runs with the new universal primer and its uncovered SSU rRNA gene produce a new set of universal primers, ultimately maximizing coverage of the target microorganism by the universal primers. On the author’s Mac M2 computer, the evaluation of primers using the Silva database generally takes around 3 min, searching and downloading genes can be completed within 2 min, and the script execution time is approximately 5 min per run, making the entire process take about 10 min. A screening recording of the script’s operation is placed in the Supplementary material. The script has been uploaded to GitHub and is available at https://github.com/haojunsp/script.git.

All eight universal primers (BA-515F-806R, BA-341F-806R, B-341F-806R, A-341F-1059R, A-784F-1059R, B-27F-1492R, BAE-515F-926R, E-616F-1132R) could be improved for targeting specific microorganisms using the “Degenerate primer 111” script. The study obtained 29 personalized primer pairs, which increased coverage of target microorganisms as well as taxa within the same domain, without affecting unrelated domains. The primer sequences and coverage changes are presented in Table 1 and Supplementary Table S2, and the script run data is available in Supplementary material.

In most cases, improved primers were obtained after 1–2 iterations. For instance, improving primer BA-515F-806R targeting Dehalococcoides. In the Silva database, there are a total of 38 16S rRNA gene sequences for Dehalococcoides, out of which 36 are not covered by the primer 515F-806R. One of the Dehalococcoides 16S rRNA genes, which was randomly chosen from those not covered by 515F-806R, was compared to the primer 515F-806R. Forward primer comparison result showed no difference in bases between the target 16S rRNA gene of Dehalococcoides and the 515F primer (GTGYCAGCMGCCGCGGTAA), while reverse primer comparison results revealed one mismatch between the Dehalococcoides 16S rRNA gene (GGACTACCAGAGTATCTAAT) and the 806R primer (GGACTACNVGGGTWTCTAAT), specifically, a G-to-A mismatch. The degenerate base for this mismatch was assigned as R, resulting in the new reverse primer sequence GGACTACNVGRGTWTCTAAT. The script running interface was shown in Figure 3. Subsequent iterations revealed that three Dehalococcoides 16S rRNA genes remained uncovered by the new primer. Two of them had two different bases compared to the previous new R primer, and one had more than three differences from the 515F primer, rendering it ineffective to modify them further due to excessive degenerate bases. Consequently, the previous effective sequence was determined as the final improved primer. The new R primer, in combination with the 515F primer, formed the new primer (515F-806R)-M. Evaluation with the Silva database showed an increase in the coverage of Dehalococcoides from 5.3 to 92.1%. The coverage of 421 other bacterial taxa also increased to varying degrees, resulting in an overall bacterial coverage increase from 83.6 to 83.8%, as shown in Figure 4. The coverage of Archaea remained unchanged at 83.5%. Meanwhile, the coverage for eukaryotes remained at 0.1%.

There were cases in this study where universal primers could not be improved. When attempting to improve primer B-27F-1492R for p_10bav-F6, p_Apal-E12, p_Fervidibacteria, p_MAT-CR-M4-B07, and p_TX1A-33, the sequences provided by Silva were shorter than the target primer length, resulting in unsuccessful improvements.

The study also involves modifying the initial primers to adapt to primer improvements. When improving primer E-616F-1132R for p_Excavata, it was found that F primer could only accommodate an additional single degenerate base, while the actual situation required two. Luckily, one of the degenerate bases was at the terminal position, so it was removed to ensure that the number of degenerate bases did not exceed the limit allowed by Silva. Researchers could employ other flexible methods, such as reducing the degeneracy of the original primer to make room for introducing new degenerate bases.

In most cases, improved primers can detect a greater variety of bacteria within the same sample. This study compared the difference in the number of microbial species detected using unimproved primers BA-515F-806R and improved primers BA-515F-806R-M1 in eight samples. In seven samples (S1, S2, SD1, SD2, SC1, SC2, SDC2), the improved primers detected more species per sequencing depth than the unimproved ones, with increases of 1.3-fold, 1.2-fold, 1.1-fold, 1.1-fold, 1.2-fold, 4.6-fold, and 1.2-fold, respectively. In one sample (SDC1), the improved primers detected fewer species per sequencing depth compared to the unimproved ones, at 0.7-fold (Figure 5B). This study compared the difference in the number of species detected at the genus level by improved and unimproved primers for 10 taxa with increased coverage by BA-515F-806R-M1 in the same sample. This effectively tested the improvement of primers using 80 actual samples. In total, 41 samples showed no detection with either primer, while 39 samples were detected with at least one primer, with 31 samples showing detection of more species with the improved primer BA-515F-806R-M1 and 8 samples showing detection of fewer species with the improved primer. Specifically, among samples where the coverage of g_Dehalococcoides increased from 5 to 92%, all 8 samples showed increased detection. Among samples where the coverage of g_Inhella increased from 90.5 to 95.2%, all 2 samples showed increased detection. Among samples where the coverage of g_Methylomicrobium increased from 93.5 to 96.8%, 4 samples showed increase, 2 showed decrease. Among samples where the coverage of g_Caminicella increased from 71.8 to 74.4%, 3 samples showed increase, 1 showed decrease. Among samples where the coverage of g_Myroides increased from 88.8 to 89.9%, all 1 sample showed increased detection. Among samples where the coverage of g_Dokdonella increased from 89.5 to 90.5%, 4 samples showed increase, 1 showed decrease. Among samples where the coverage of g_Desulfovibrio increased from 88.5 to 89.4%, 5 samples showed increase, 2 showed decrease. Among samples where the coverage of g_Pedomicrobium increased from 88.3 to 89.0%, 4 samples showed increase, 2 showed decrease. Among samples where the coverage of g_Lewinella increased from 91.9 to 92.6%, all 4 samples showed increased detection. Among samples where the coverage of g_Turicibacter increased from 81.3 to 81.9%, all 4 samples showed increased detection (Figure 5A).