Protocol for screening and expression studies of T-DNA and tagging-based insertional knox mutants in Arabidopsis thaliana

Abstract

KNOTTED1-like homeobox (KNOX) genes serve important roles in meristem function and many developmental processes in all higher plants. In Arabidopsis, studies of KNOX genes especially among members of class II KNOX genes remain limited and functional data are largely lacking. In the present study, we established a reproducible protocol that is important for genetic studies of KNOX genes using Arabidopsis insertional mutants. This protocol contains a reproducible and serial procedure containing detailed and step-by-step laboratory and field works covering all experiment steps from the screening of homozygous mutant lines to the KNOX expression analysis using qRT-PCR in a single paper. The troubleshooting and challenges that might occur are also presented and discussed. T-DNA insertion mutants for all Arabidopsis KNOX genes (except for knat4) were isolated based on kanamycin screening, phenotype selection, and PCR genotyping. Surprisingly, the insertions resulted in strong repression of the respective KNOX genes. However, no gene suppression was observed for the positively selected knat5 mutant. Moreover, qRT-PCR was effective for transcript analysis among the knox mutant samples. The use of different relative expression quantification produces a similar indication of expression level. Overall, the proposed procedure is highly effective for expression studies of KNOX genes in Arabidopsis mutants and will serve as a fundamental work protocol to open opportunities for genetic studies of genes involving insertional mutants in Arabidopsis.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-021-02868-8.

Introduction

Functional gene analysis in Arabidopsis thaliana has greatly contributed to functional studies of the genes responsible for many biological traits in Arabidopsis and other plants (Rhee and Mutwil 2014). Arabidopsis was developed as the first model plant due to its short life cycle and small genome size of only 10 chromosomes (2n = 10). Since its genomic DNA has been completely sequenced (The Arabidopsis Initiative 2000), the functional study of its genes has become a major research topic in many molecular biology laboratories. This data has sped up investigations of many biological and biochemical contributions of the genes. However, functional studies of the Arabidopsis genome remain incomplete.

The use of mutants has proven to be a beneficial strategy for studying specific gene functions, especially in A. thaliana. Notably, the production of mutants in Arabidopsis is more feasible when compared to other plants (especially tree species) since the generation of mutants is typically laborious, ineffective, and unspecific. One direct method of measuring a gene’s functions and contributions to biochemical, cellular, tissue, and organ characteristics involves the analysis of mutants (O’Malley et al. 2015). However, elucidating gene function requires careful analyses of the phenotypes caused by mutations. One technique developed to effectively produce Arabidopsis mutants is transfer DNA (T-DNA) insertional mutagenesis, which has had a major impact on plant molecular biology research (Fieldman 1991; Koncz et al. 1992). Arabidopsis T-DNA insertional mutants are derived from the Agrobacterium-mediated transformation of T-DNA, which is a specific DNA segment that is transferred from Agrobacterium into Arabidopsis plant cells. T-DNA insertions cause the inactivation or alteration of gene expression, which results in “loss of function” phenotypes (Koncz et al. 1987). Publicly available information regarding Arabidopsis mutants can be accessed in The Arabidopsis Information Resource (TAIR) at www.Arabidopsis.org.

Although T-DNA insertional mutants are abundantly available, the identification of individual mutant seeds for a certain mutated gene remains labor-intensive and involves in vitro screening, DNA genotyping, and even the classical genetic approach. The availability of gene sequences and their transcripts in TAIR makes it possible to perform molecular-based analyses such as reverse genetics when screening mutants for specific genes and elucidating their potential impact on biological functions (Bolle et al. 2011). In practice, the genetic study of a gene is conducted by measuring the gene expression of its mutants at both the transcript and protein levels compared to the wild type. For the present study, laboratory work involving the homozygous identification of T-DNA insertional mutants and molecular analysis of their expression was performed. Since some problematic cases can be found during experiments, an established and reproducible protocol should be generated.

Most functional studies of Arabidopsis KNOTTED1-like homeobox (KNOX) genes have benefited from insertional mutants. KNOX genes are families of homeobox genes that have been identified in all monocot and dicot species, while subsets of these genes regulate meristem function in all higher plants (Scofield and Murray 2006). Homeobox genes encode proteins containing a conserved DNA-binding homeodomain motif found in transcription factors from all eukaryotes. Most homeobox genes encode transcription factors, which function in developmental processes. The eight KNOX genes in Arabidopsis are divided into two subfamilies. The subfamily KNOX I comprises STM, KNAT1, KNAT2, and KNAT6, while and the subfamily KNOX II comprises KNAT3, KNAT4, KNAT5, and KNA7 (Scofield and Murray 2006). Gene expression analysis using knox mutants in Arabidopsis has been established many years ago, but a reproducible protocol containing a serial procedure with detailed and step-by-step laboratory and field works covering all experiment steps from the screening of homozygous mutant lines to the KNOX expression analysis using qRT-PCR has not been reported yet. In the present study, we report all required experiments for the expression studies of KNOX genes using Arabidopsis insertional mutants in a single paper.

Here, we report a comprehensive, established, and reproducible protocol that is important for the genetic study of KNOX genes using Arabidopsis insertional mutants. The troubleshooting and challenges that might occur in this process are also presented and discussed.

Materials and methods

Plant materials

The Arabidopsis lines (i.e., both the knox mutants and wild type) were obtained from the Nottingham Arabidopsis Stock Center (NASC; http://nasc.nott.ac.uk). Seeds of knat1^bp−9 were provided by Dr. A. Hay (Department of Plant Sciences, Oxford University, UK). A list of plant materials used in this experiment is presented in Table 1.

Table 1.

Arabidopsis knox mutants and wild type used in the present study

No	NASC stock number/source	Name of line/mutant	Insertion/mutation	Ecotype
1	N28166	Columbia-0 (Col-0)/ Wild type	–	–
2	Dr. Angela Hay, Oxford University	knat1bp−9	dSpm transposon, 1st intron	Col-0
3	N409575	stm-GK	T-DNA, 1st intron	Col-0
4	N609159	knat2	T-DNA, 3rd intron	Col-0
5	N636464	knat3	T-DNA, 1st intron	Col-0
6	N520216	knat4	T-DNA, 1st intron	Col-0
7	N616798	knat5	T-DNA, 1st intron	Col-0
8	N617904	knat6	T-DNA, 3rd intron	Col-0
9	N610899	knat7	T-DNA, 2nd intron	Col-0

In vitro seed screening

To identify homozygous knox mutants carrying T-DNA insertions, a modified version of the method from Feldmann (1991) was used for kanamycin selection. Seeds were surface sterilized using 70% ethanol for 1 min and 5% calcium hypochlorite for 30–40 min. The seeds were then washed three times using sterile water. The sterilized seeds were then stored for 3–5 days at 4 °C for seed stratification. Thereafter, the seeds were placed on selective agar plates containing 1/2 Murishage and Skoog (MS) nutrients, 35 µg/ml kanamycin (Sigma, Steinheim, Germany), 1% (w/v) solidified agar, and 1% (w/v) sucrose (pH 5.8). The seeds were then arranged in a single horizontal row on 10 cm² Petri dishes. The plates were wrapped with parafilm and incubated at 25 °C in a growth room [constant fluorescent light (500–750 lx); 25 °C]. The plates were placed in a vertical position to allow the roots from most plants to grow along the surface of the agar. Fifty seeds were inoculated to the kanamycin screening medium for each Arabidopsis knox mutant and the wild type. The kanamycin-resistant seedlings, which had greener cotyledons and longer primary roots, were observed and the number of resistant or sensitive seeds was counted. The resistant seeds were transferred to soil and grown in long-day conditions (16 h light, 8 h dark). The plants were fertilized with 1/2 MS nutrients at the first week after planting. The plants were maintained in the soil and the leaves were subsequently used for mutant genotyping. The seeds were harvested and stored for expression analysis or a future investigation.

Mutant genotyping

The leaves of resistant plants growing in soil were collected for genomic DNA extraction. DNA was extracted using a simple and rapid method (Edwards et al. 1991). PCR genotyping was then performed with specific primers designed against the T-DNA insertion and wild-type allele. Primer design was performed at http://signal.edu/tdnaprimers.2.html. The primer positions and T-DNA insertion location of a prospective KNOX gene are presented in Fig. 1A. The list of primers used for mutant genotyping is presented in Table 2a.

Fig. 1.

a Primer construction for homozygous mutant identification using PCR genotyping. LP left primer, RP right primer, and LBa1: the primer for T-DNA insertion. b Location of flanking sequences and insertions in the knox mutant genes. Flanking sequences are indicated by dashed-boxes and introns by black triangles. The insertional fragments are T-DNA or transposons. a–j ACTIN2 and knox mutant genes

Table 2.

Primer pairs used for knox mutant PCR genotyping (a) and expression analysis using qRT-PCR (b)

SALK number	Locus	Oligo name	Sequences	Length (bp)	Expected product size (bp)
a. PCR genotyping
SALK_109159	AT1G70510	knat2 (LP)	GAGTTTGTCCTTGCCTTCATG	21	–
SALK_109159	AT1G70510	knat2 (RP)	TCCAGCTAGTTCTTATCAGGTGG	23	–
SALK_136464	AT5G25220	knat3(LP)	TCTCCTTCAATCATTTCACCG	21	–
SALK_136464	AT5G25220	knat3 (RP)	ACATCTAATCCCCCATCGAAC	21	–
SALK_116798	AT4G32040	knat5 (LP)	TTCGGAGATGCAAAATACTGG	21	–
SALK_116798	AT4G32040	knat5 (RP)	TTGATGTACCATTGGAGCTTG	21	–
SALK_117904	AT1G23380	knat6 (LP)	TTATCCCTCTCTGGTTCGGTC	21	–
SALK_117904	AT1G23380	knat6 (RP)	GCAGATAAGAGTGGCCACTTG	21	–
SALK_110899	AT1G62990	knat7 (LP)	TTGCCACCAATT TTTCAAGAC	21	–
SALK_110899	AT1G62990	knat7 (RP)	TGCCGTGAAATTGAGAACAAC	21	–
	T-DNA	LBa1	TTGTTCACGTAGTGGGCCATC	21	–
b. Expression analysis using qRT-PCR
–	AT3G18780	Actin2 (LP)	tgggatgaaccagaaggatg	20	60
–	AT3G18780	Actin2 (RP)	aagaatacctctcttggattgtgc	24
–	AT1G62360	STM1 (LP)	tcctcaccttcctctttctcc	21	142
–	AT1G62360	STM1 (RP)	gcaagagctgtcctttaagctc	22
–	AT1G62360	STM2 (LP)	caaatggccttacccttcg	19	104
–	AT1G62360	STM2 (RP)	gccgtttcctctggtttatg	20
–	AT4G08150	KNAT1 (LP)	tcccattcacatcctcaaca	20	86
–	AT4G08150	KNAT1 (RP)	cccctccgctgttattctct	20
–	AT1G70510	KNAT2 (LP)	cagcgtctgctacagctcttt	21	85
–	AT1G70510	KNAT2 (RP)	tcatccgctgctatgtcatc	20
–	AT5G25220	KNAT3 (LP)	gaagaacaaacgcaaaaggtg	21	113
–	AT5G25220	KNAT3 (RP)	ctaaaaccctgctttcaaatcc	22
–	AT4G32040	KNAT4 (LP)	cagtcgcttcaaagttttacagg	23	78
–	AT4G32040	KNAT4 (RP)	ttgctcatcttcatcctcagac	22
–	AT1G23380	KNAT5 (LP)	aatggccatacccaactgag	20	128
–	AT1G23380	KNAT5 (RP)	tgacgtggaagagttgctgt	20
–	AT1G62990	KNAT6 (LP)	gtctgccaggggagtttct	19	75
–	AT1G62990	KNAT6 (RP)	gctacctcatgatcacctcca	21
–	AT1G70510	KNAT7 (LP)	ttgccgtgaaattgagaaca	20	90
–	AT1G70510	KNAT7 (RP)	tcatcctcatcctccgacat	20

PCR was run for 35 cycles. This process involved denaturation at 94 °C for 40 s, annealing at 55 °C for 40 s, and extension at 72 °C for 1 min and 20 s. The PCR products were loaded on a 1% agarose electrophoresis gel and run at 70–100 V for 20–30 min. PCR reagents and procedures were provided by Fermentas (St. Leon-Rot, Germany). The PCR amplification products were determined and run on 1% polymerized agarose in 1 × Tris–acetate-EDTA (TAE) buffer for 40 min at 100 V. DNA bands were stained with RedSafe™ and visualized on a UV transilluminator. The gel documentation system of Bio-Rad, USA was applied to capture the gel electrophoresis profile. To measure the size of DNA bands, the O’GeneRuler Express DNA Ladder (Thermo Scientific, USA) was employed as a marker. The emergence of a discrete band in amplifications using only the LBa1 and RP primers indicated a homozygous line. The seeds of homozygous plants were collected and separated for the expression analysis.

Phenotypic screening

Phenotypic screening was performed for knat1^bp−9 and stm-GK since the mutant phenotypes of both lines are well characterized (Venglat et al. 2002; Barton and Poethig 1993) and only appeared in the homozygous lines. The seeds of knat1^bp−9 and stm-GK mutants were grown on soil and maintained in long-day conditions for homozygous line selection. The selection of homozygous knat1^bp−9 was based on the phenotype of downward-pointing siliques (Venglat et al. 2002). Meanwhile, the selection of stm-GK was based on the delayed initiation of a functional shoot apical meristem (SAM) (Barton and Poethig 1993; Long et al. 1996; Clark et al. 1996).

Expression analysis

Total RNA extraction and cDNA generation

Fresh tissue samples collected from selected 6-week-old homozygous and wild-type plants were immediately dipped in liquid nitrogen after slicing and stored at − 80 °C until the total RNA extraction were performed. Total RNA was isolated from fresh hypocotyls or other tissues (leaf, inflorescence, flower, and pedicel) using the RNeasy Plant Mini Kit (Qiagen, Hilden, Germany) and following the manufacturer’s guidelines. RNA was quantified via UV spectrophotometry and the quality was tested by running an aliquot on a 1.5% agarose gel. A typical gel of total RNA extraction represented by 28 s and 18 s rRNA in distinct bands with a smear of mRNA between and above those bands was observed. Reverse transcription (RT) was performed to generate the first-strand cDNA from total RNA using Quantitec^® Reverse Transcription (Qiagen) according to the manufacturer’s guidelines, which include gDNA digestion before the RT reaction. Thereafter, 1 µg of total RNA was used for first-strand cDNA synthesis. cDNAs were PCR-amplified using an endogenous reference gene primer pair (i.e., ACTIN2, a housekeeping gene) and run on a 1.5% agarose gel to evaluate the quality of the first-strand cDNA. PCR amplification was run for 30 cycles, with denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 1 min. The PCR reaction procedures were provided by Fermentas (St. Leon-Rot, Germany) and involved using 0.5 µl cDNA instead of a 2 µl DNA template (as per qRT-PCR).

Primer design

The Universal ProbeLibrary Design Center (http://www.roche-applied-science.com/sis/rtpcr/upl/ezhome.html) was used to design specific primers for qRT-PCR. Primers were designed to amplify flanking sequences located downstream of or spanning the T-DNA/transposon insertion while spanning an intron, if possible. This design allows researchers to study the impact of an insertion on its own insertion site as well as classical expression analysis. The location of flanking sequences and insertions in KNOX genes are shown in Fig. 1B. A list of the primers used for qRT-PCR is presented in Table 2b.

To evaluate the designed primers, the cDNAs of the wild type (Col-0) were PCR-amplified using the primers and run on a 1.5% agarose gel. The aforementioned PCR amplification procedure applied to the ACTIN2 primer pair for testing the first-strand cDNA quality was employed for this purpose.

qRT-PCR

qRT-PCR was performed using the Roche qRT-PCR SYBR green kit (Roche, Grenzach-Wyhlen, Germany) and reactions were run on a LightCycler^®480 (Roche, Grenzach-Wyhlen, Germany) according to the manufacturer’s protocol: preincubation at 95 °C for 5 min, amplification at 95 °C for 10 s, 61 °C for 10 s, 72 °C for 10 s, melting curve at 95 °C for 5 s, 65 °C for 60 s, 67 °C—Acqu. 5, and cooling to 40 °C.

Data analysis and folding expression calculation

Data were analyzed using LightCycler^®480 Software v1.5.0 (Roche Grenzach-Wyhlen, Germany). Values for crossing points (CP) were directly obtained from the software and transformed to absolute concentration values using the following formula:

$$\text{CP}=\text{Slope}.\text{log}\left[x\right]+Y\text{intercept, which}x={\left(10\right)}^{\left[\frac{\text{CP}-Y\text{intercept}}{\text{Slope}}\right]}$$

where CP is a crossing point value, (x) is the concentration of amplified cDNA at time point 0, and slope and Y-intercept) are the values obtained from running standard curves generated by template dilution.

The absolute concentration values (x) were then normalized to the expression of ACTIN2 by dividing the absolute expression value of the gene of interest by the absolute expression value of ACTIN2 in the corresponding samples. All experiments were performed using three biological and three technical replicates unless otherwise stated. To determine the slope (efficiency) and intercept, standard curves from the dilution series were calculated. The initial first-strand cDNA (1 µg of total RNA) was diluted 5×, corresponding to standard 1. Then, a series of 5 × dilutions starting from standard 1 was made. For the standard curve, dilutions from 5⁰ to 5^–7 were used. This procedure was performed for all employed primer pairs. The simple mathematical model for relative expression quantification from Pfaffl (2001) (i.e., $\text{Expression Ratio}\left(R\right)=\frac{{\left(E\text{target}\right)}^{\text{CP target}\left(\text{Control}-\text{Sample}\right)}}{{\left(E\text{ref}\right)}^{\text{CP ref}\left(\text{Control}-\text{Sample}\right)}}$) was also employed as a comparison, where E_target is the real-time PCR efficiency of the target gene transcript, E_ref is the real-time PCR efficiency of the reference gene transcript, ΔCP_target is the CP deviation of the control—sample of the target gene transcript, and ΔCP_ref is the CP deviation of the control—sample of the reference gene transcript.

Results and discussion

Homozygous KNOX mutant screening

Functional studies of genes using gene-tagging mutants in Arabidopsis must employ a homozygous line since it carries all mutated alleles and its selfing seeds will not be phenotypically segregated. The use of a homozygous line ensures that the phenotype is controlled by a gene of interest. However, in some cases, the effect of gene-tagging mutation varies in mutant lines, which is shown by differences in phenotype severity among alleles. Since mutant alleles might have different strengths in terms of their mutant severity, the experiment used the appropriate mutant alleles.

Seed propagation represents important initial work that must be performed after receiving seeds from a mutant seed stock center such as the NASC. The number of seeds provided is usually limited (20–100 seeds) and these seeds might also contain a small proportion of segregating or homozygous lines. The direct use of seeds is not recommended since it increases the risk of seed sufficiency if some seeds are no longer viable or a technical problem occurs during the kanamycin selection or mutant genotyping processes. For Arabidopsis, seed propagation is commonly conducted by sowing seeds in soil and growing them under long-day conditions (see “Materials and Methods”).

Kanamycin selection differentiated the mutant seeds into two groups: sensitive seedlings (Fig. 2a) and resistant seedlings (Fig. 2b). Resistant seedlings exhibited green cotyledons and long roots, while sensitive seedlings exhibited a pale color and inhibited growth. All wild type (Col-0) seedlings were sensitive (Fig. 2a) and all kanamycin-resistant seedlings were homozygous or heterozygous (Fig. 2b, c). All seedlings that were resistant to kanamycin (Fig. 2b) were derived from a homozygous mutant line. Additionally, seeds derived from a heterozygous mutant line became segregating kanamycin-resistant seedlings (Fig. 2c).

Fig. 2.

Kanamycin screening of homozygous knox mutants. a–c Seedlings growing on kanamycin selective medium, a sensitive seedlings (wild type, Col-0); b resistant seedlings derived from a homozygous mutant line; c segregating resistant seedlings derived from a heterozygous mutant line

The kanamycin selection of knox mutants showed a variety of responses (Table 3). Resistant seedlings identified via kanamycin selection were homozygous or heterozygous mutants. Putative homozygous mutants were observed at knat3 and knat4 since all seedlings were resistant and showed no segregation in the kanamycin screening. Notably, DNA genotyping is required for further verification to select the homozygous lines. While the direct genotyping of grown primary seeds instead of kanamycin selection is also possible, kanamycin selection reduces negative mutants and the sample number for DNA genotyping. However, expression of the nptII (kanamycin-resistant) gene might involve silencing and the T-DNA transgenic Arabidopsis could thus not survive in the kanamycin medium (Francis and Spiker 2005). Therefore, it is also possible to conduct direct DNA genotyping without an initial kanamycin screening (O’Malley et al. 2015).

Table 3.

Kanamycin selection of knox mutants

No	Arabidopsis mutant*	SALK line number/mutants number	Number of inoculated seeds	Number of germinated seeds	Number of resistant seedlings	Number of sensitive seedlings	Mutant genotype origin
1	Col-0	–	50	25	0	25	Wild type
2	Stm-GK	GABI_100F11	49	48	32	16	Heterozygous
3	knat2	SALK_099837	48	46	29	17	Heterozygous
4	knat3	SALK_136464	45	45	45	0	Homozygous
5	knat4	SALK_020216	46	46	46	0	Homozygous
6	knat5	SALK_116798	50	50	39	11	Heterozygous
7	knat6	SALK_117904	49	49	36	13	Heterozygous
8	knat7	SALK_110899	50	44	32	12	Heterozygous

The knat1^bp−9 mutant was not included in the screening since it is not a T-DNA insertional mutant

Phenotype-based homozygous line selection among stm and knat1 mutants

Since the phenotypic characteristics of the stm and knat1/BP mutants can be visually observed during their growth, the selection of homozygous lines can be performed using their phenotypes (Fig. 3). By sowing heterozygous seeds, the phenotypes of seedlings can be observed as a segregated line due to the Mendelian rule dictating that only one-fourth of the seedling population should exhibit the homozygous stm or knat1/bp mutant phenotypes (recessive homozygous). STM is responsible for maintaining the meristematic cell pool in the SAM. The homozygous line exhibits an inactivated STM gene, which results in the delayed initiation of a functional SAM due to the lack of an embryonic shoot meristem (Barton and Poethig 1993), as shown in Fig. 3g. Since the weak allele of the stm mutant was used in this experiment, the adventitious meristems continued to grow (Fig. 3h) and produce non-functional inflorescences (Fig. 3i). Heterozygous plants exhibited a normal phenotype that is normally observed in the recessive homozygous genotype. Therefore, maintaining heterozygous seeds is very important for future experiments since homozygous plants cannot produce seeds (sterile). Furthermore, the strong allele of the stm mutant shows fused cotyledons and the cessation of meristematic growth; thus, no further growth is observed (Barton and Poethig 1993). For genetic studies, the use of mild or weak alleles is especially recommended if the study aims to investigate the effect of a gene in advancing organ formation (e.g., inflorescence, stem and leaf formation, etc.). Moreover, KNAT1 or BREVIPEDICELLUS is a key regulator of inflorescence architecture in Arabidopsis (Venglat et al. 2002). The mutation of this gene (e.g., knat1^bp−9 in this study) produces downward-pointing siliques (Fig. 3f) that were not observed in the wild type (Fig. 3c). Since the regenerative organs of the homozygous knat1^bp−9 are functional (Fig. 3e, f), the inherited seeds (T2) of the homozygous mutants are also homozygous. In genetic studies, homozygous mutant selection benefits greatly from these specific phenotypic characteristics.

Fig. 3.

Homozygous mutant phenotypes from segregating seedlings of stm-GK and knat1^bp−9 mutants in Arabidopsis. a–c Seedling, initial flowering, and inflorescence of the wild type (Col-0); d–f seedling, initial flowering, and downward-pointing siliques of the knat1^bp−9 mutant; g delayed initiation of a functional shoot apical meristem (SAM) by the stm-GK mutant; h adventitious growing shoot; i non-functional inflorescence (producing sterile flowers). Red arrows indicate downward-pointing siliques at initial flowering (e) and for mature plants (f)

DNA genotyping of knox mutants

A simple and rapid DNA extraction method (Edwards et al. 1991) is sufficient for knox mutant genotyping using a leaf sample from an actively growing Arabidopsis plant. DNA quantification using a spectrophotometer is important to ensure a sufficient amount of DNA (shown by concentration) and DNA quality (value of λ 260/280 between 1.8 and 2.0). As an additional test of DNA quality, running DNA in an agarose gel can be beneficial for resolving issues in mutant genotyping especially if there is a problem in DNA amplification. Another comprehensive DNA extraction procedure is also recommended; however, it is time consuming and laborious when mutant genotyping using many samples. The specific flanking sequence at the knox mutant has been successfully amplified using DNA primers generated from the iSect tools for the SALK line (http://signal.salk.edu/tdnaprimers.2.html) (Fig. 4). The detailed information on designing primers of the SALK line is provided by O'Malley et al. (2015). The PCR genotyping of knat2, knat3, knat5, knat6, and knat7 was successfully performed, while it failed for knat4 (Fig. 4). Although all seedlings were resistant to the kanamycin selection medium (indicating a homozygous line, Table 3), no amplicon was detected in the gel during knat4 genotyping (data not shown). This might be a false positive line that does not contain the T-DNA at the identified locus. Notably, such lines are found in 12.6% of SALK lines (O'Malley et al. 2015). In the homozygous lines, flanking sequences were amplified using only Lba1-RP primers, while amplification was only detected when using LP-RP primers for the wild type (Fig. 4). Additionally, the flanking sequences were amplified using both Lba1-RP and LP-RP primers for the heterozygous lines. The genotyping results indicate that all selected knox3 plants were homozygous mutants, while knox2, knox5, knox6, and knox7 were shown as segregating lines (i.e., heterozygous mutants) (Fig. 4). These results are in line with the kanamycin selection results (Table 3). The band thickness variation detected by gel documentation (Fig. 4) highlights the variation in DNA concentrations during DNA extraction. Therefore, DNA concentration adjustment among lines in the PCR genotyping is suggested to increase the homogeneity of band performance, which was not applied in this study.

Fig. 4.

A typical gel for PCR genotyping of the wild type shows a discrete band of approximately 1000–1500 bp when amplified by LP/RP primers. No amplicon was detected for the wild type when amplified using the Lba1/RP primer combination. In plants, homo- or heterozygous amplification with Lba1/RP resulted in a specific band of 500–900 bp. a knat2 genotyping, with lines 1, 2, 4, and 5 homozygous for the insertion; b knat3, all lines tested homozygous; c knat5, all lines tested homozygous except for lines 8 and 9; d knat6, lines 1, 2, 4, 6, 8, and 10 were homozygous; e knat7, only line 10 was homozygous. Marker (M): 1 kb DNA ladder (Fermentas, Germany)

KNOX expression study

Total RNA was extracted from fresh tissue, depending on the tissue-specific interest of gene expression. In this experiment, RNA was extracted from hypocotyls, leaves, flowers, pedicels, first node inflorescences, and the oldest node inflorescences. The concentration-adjusted RNA from the hypocotyls of the knox mutants was loaded in the agarose gel (Fig. 5). Typical distinct bands representing 28 s and 18 s rRNA appeared (Fig. 5). The smear bands above and between the two bands showed all mRNAs. The two distinct bands and transparent smear bands that appeared in the gel (Fig. 5) suggest the good quality of extracted total RNA and RNA concentration adjustment accuracy. If the bands show a lower or lack of signal for the distinct bands, it is recommended that the researcher repeats the extraction since it indicates the low quality of the total RNA.

Fig. 5.

A typical gel for total RNA extraction. Distinct bands representing 28 s and 18 s rRNA. M: 1 kb DNA ladder. (Lanes 1–8) Total RNA of stm-GK, knat1^bp−9, knat2, knat3, knat4, knat5, knat6, and knat7 mutants

RT was performed to produce the cDNA from the total RNA. cDNA quality was evaluated by amplifying an endogenous standard (i.e., a reference housekeeping gene primer pair (ACTIN2)) and running it on an agarose gel (Fig. 6a). The use of a housekeeping gene is suggested for cDNA evaluation because the gene is constitutively expressed in all Arabidopsis cells. Distinct homogenous bands of the same size (60 bp) indicate the good quality of the cDNA. In this study, cDNA generation and qRT-PCR were conducted in separate and serial experimental steps (two-step RT-PCR). A one-step RT-PCR process has been developed in which RT and RT-PCR are performed sequentially via a single reaction within the same tube. Although one-step RT-PCR shortens the RT-PCR procedures and takes less time than the two-step procedures, the separated procedures have the same benefits (i.e., enabling the evaluation of cDNA quality before the RT-PCR process to ensure successful expression during qRT-PCR).

Fig. 6.

a PCR amplification of ACTIN2 from the cDNAs of stm-GK, knat1^bp−9, knat2, knat3, knat4, knat5, knat6, and knat7 mutants (lanes 1–8). b PCR amplification of the wild type (Col-0) cDNAs using ACTIN2 (lane 1) and KNOX primer pairs, i.e., STM1, STM2, KNAT1, KNAT2, KNAT3, KNAT4, KNAT5, KNAT6, and KNAT7 (lanes 2–10). M: 1 kb DNA ladder

All designed KNOX gene primers effectively amplified the cDNAs generated from the hypocotyl RNA of the Arabidopsis wild type (Fig. 6b). All bands produced were specific at sizes that varied from 60 to 128 bp, as expected (Table 2b; Fig. 6b). A quality test for the primers was conducted to evaluate the specificity of RT-PCR products and resulted in a single product with the desired length. The two primers of the STM gene (i.e., STM1 and STM2) used in this study were subjected to a primer positional test of the flanking sequences toward the insertion position. The flanking sequence of the STM1 primer pair is located downstream of or spanning the T-DNA insertion; on the other hand, the flanking sequence of the STM2 primer pair is upstream of the T-DNA insertion (Fig. 1b). The power of the two STM primer pairs will be studied in the respective mutant since the success of expression studies relies on the selection of good primers that flank in the appropriate locations toward insertions to result in the strong downregulation of genes.

For plants, transcript analyses provide the comparative analysis of spatial and temporal gene expression for various transcripts among samples (Geniza and Jaiswal 2017). qRT-PCR detected the expressed genes and abundance of transcripts by amplifying the existing cDNAs using a specific gene primer pair to produce raw data in the form of CP values. Using the given formula, the data for relative expression were calculated by normalizing to the ACTIN2 expression (Supplemental data). A simple statistical analysis, such as a t test, can compare the statistical differences in relative expression level (Supplemental data). Alternatively, column graphs or other types of graphs could be applied to provide a better interpretation of differences in gene expression (Fig. 7). The normalized expression data of the samples (tissues) by a reference gene (ACTIN2) was sufficient for the comparative analysis of tissue-specific expression since no specific tissue acts as a control for the relative quantification. Alternatively, we could choose one of the tissues/body parts as a control based on the specific criteria (e.g., the highest expression level) and then calculate the relative expression level of the target sample with the control (wild type). In the present study, a simple statistical analysis (t test) was employed by comparing the relative expression at each specific tissue with a selected control (hypocotyl) (Supplemental data). Alternatively, a multiple mean analysis could also be used with no control sample.

Fig. 7.

Tissue-specific expression of Arabidopsis STM, KNAT1, and KNAT7 in the wild type (Col-0) of 6-week-old plants

The differential expression of KNOX genes was observed among Arabidopsis tissues since STM and KNAT1 were strongly expressed in flowers and hypocotyls and very weak in leaves (Fig. 7). However, the highest expression was detected in young floral apices containing floral meristems and all floral stages until pollination. Notably, expression in leaves was low. In contrast, KNAT7 expression was highest in the oldest node inflorescence, while low expression was detected in hypocotyls. A similar expression pattern observed for STM and KNAT1 (Fig. 7) indicates that these two genes’ tissues may have a redundant function and the relatively high expression in hypocotyls.

Gene expression analysis using 6-week-old hypocotyls showed that the tagged KNOX genes were dramatically downregulated in stm-GK, knat1^bp−9, knat2, knat3, knat6, and knat7, respectively (Table 4). However, the magnitude of this reduction varied for the different mutants. Analysis of the expression level of respective mutants compared to the wild type was important to observe since it ensures that the mutation is functional. The insertional mutations at stm-GK, knat1 ^bp−9, knat2, knat3, knat6, and knat7 were functional and could strongly downregulate the respective gene expression, especially at the Arabidopsis hypocotyl (Table 4). Conversely, the insertions could not repress KNAT5 expression in knat5 mutant, even if they were upregulated. However, knat5 was positively selected at kanamycin screening and PCR genotyping for homozygous mutant selection (Table 3; Fig. 4); thus, the T-DNA insertion cannot suppress KNAT5 expression and the transcripts are still produced. In the case of knat5, the elevated gene expression observed might be due to insertion into a negative gene regulatory element or the increased stability of hnRNA.

Table 4.

Relative expression level* and the ratio of fold change** in KNOX gene expression for mutants compared to the wild type (Col-0) based on hypocotyl tissue samples

Lines	X	SD	p	Fold change (times)***	Fold change (times)****	Expression
STM1
Col-0	3.4507	0.3456
stm-GK	0.0046	0.0078	0.0017	– 745.70	– 6652.87	Downregulated
STM2
Col-0	1.2303	0.2129
stm-GK	5.0593	0.2826	0.0024	+ 4.11	+ 4.14	Upregulated
KNAT1
Col-0	1.3069	0.0915
knat1bp−9	0.0054	0.0032	0.0008	– 242.84	– 286.84	Downregulated
KNAT2
Col-0	2.2320	0.3455
knat2	0.0070	0.0056	0.0040	– 317.72	– 424.70	Downregulated
KNAT3
Col-0	0.0940	0.0211
knat3	0.0022	0.0005	0.0085	– 42.42	– 42.47	Downregulated
KNAT4
Col-0	0.2234	0.0589
knat4	0.2233	0.0394	0.4989	+ 1.00	+ 1.02	Downregulated
KNAT5
Col-0	1.0151	0.2136
knat5	17.1493	0.6770	0.0001	+ 16.89	+ 17.16	Upregulated
KNAT6
Col-0	0.0134	0.0031
knat6	0.0000	0.0000	0.0088	– 1967.07	– 13,276.89	Downregulated
KNAT7
Col-0	0.1866	0.0315
knat7	0.0013	0.0002	0.0047	– 139.92	– 139.40	Downregulated

X Mean value, SD standard deviation, p p value of statistical analysis using a t test compared to the wild type (Col)

*Relative concentration values were normalized to the ACTIN2 expression

**Calculated by dividing the mean of relative concentration values for knox mutants by the value for the wild type (downregulated or upregulated)

***Calculated using the formula presented in this work

****Calculated using Pfaffl’s (2001) formula. All data were calculated based on three biological and three technical replicates.

The unsuccessful screening for knat4, especially at PCR genotyping, indicates a false positive line. No gene suppression of KNAT4 at knat4, even when upregulated (Table 4), suggests that the knat4 false positive line produces regular KNAT4 transcripts. False positive events, in which T-DNA is not inserted at the identified locus, often occur in the SALK or SAIL lines. Small proportions of the SALK (12.6%) and SAIL (14.5%) collections do not contain T-DNA at the identified locus (O’Malley et al. 2015). Since the KNAT4 and KNAT5 primer pairs functioned correctly in the primer quality test (Fig. 6b), the failure in downregulation observed in both lines might be due to troubleshooting reasons. The most common reasons for this outcome include no T-DNA insertion in the line (false positive) or the lines containing t-DNA insertion producing only wild type and hemizygous plants (but no homozygous lines). Another reason could be multiple insertion events. The insertions are located in the same or nearly the same site and have similar identification numbers since they share the same insertion site. Unsuccessful PCR genotyping could be due to primer failure, where the amplified flanking sequence position does not span the T-DNA and causes the line to be genotyped as the wild type (O’Malley et al. 2015).

The different relative quantification methods performed in this study produced the same indication of expression level (Table 4). Different fold changes were detected between the quantification performed in this work and the quantification using Pfaffl’s (2001) method, especially at very low expression levels such as those observed for stm-GK (using STM1 primer pair) and knat6 (Table 4). However, these differences do not change the indication (up- or downregulation). Both quantification methods can effectively calculate the relative expression level and are recommended for relative expression quantification. Moreover, since the simple quantification method (Pfaffl 2001) needs a control as a comparison for the target samples, it less useful for the analysis of the tissue-specific gene expression especially if all samples have the same position and the control does not exist.

Furthermore, the effect of the intron insertion on the expression of the 3′ downstream sequence compared to the 5’ upstream was examined. Two different primer pairs for STM (STM1 and STM2) were designed for either downstream (STM1) or upstream (STM2) of the T-DNA insertion (Fig. 1b). The downstream flanking sequence was very poorly expressed (746 × lower in stm-GK), whereas the upstream sequence was more strongly expressed (4 × stronger in stm-GK) than in the wild type (Col-0) (Table 4). Therefore, it is strongly recommended that researchers check that the primers are designed to either span or locate 3′ downstream of the T-DNA/transposon insertion site before conducting qRT-PCR.

All of the analyzed insertions are located in introns (Fig. 1b), which may interfere with splicing or RNA stability. With the exception of knat2-5, no other exon insertions for KNOX genes have been described to date. Since T-DNA does not integrate preferentially into introns (Kim and Veena 2007), exon insertions in KNOX genes may strongly interfere with embryo or seedling viability. Alternatively, the structure of KNOX genes may hinder the integration of T-DNA. Other means of disrupting gene function include the use of chemical mutagens and RNAi (RNA-interference). While chemical mutagenesis is not applicable for species with a long generation time, RNAi has been proven as a powerful technology in various plant species (e.g., poplar). RNAi allows researchers to isolate transgenic lines with a broad range of suppression; however, target gene suppression greater than 90% is rare (Li et al. 2012). In the case of stm-GK, a suppression of over 99.8% was measured (Table 4); however, the mutant phenotype in the hypocotyl of stm-GK was very mild when compared to the null alleles (Barton and Poethig 1993). Hence, such a phenotype may not have been observed using RNAi since the degree of downregulation is generally too low. Insertional mutagenesis often leads to the expression of truncated proteins (e.g., the 5′ flanking sequence of the insertion). Such truncated proteins may still be fully functional, partially functional, or even produce dominant effects. In crossing experiments, no dominant phenotype was observed for any of the studied mutants. However, expression of the 5’ flanking sequence in stm-GK was slightly but significantly overexpressed. Since the insertion in stm-GK is in the second intron, overexpression of the 5′ flanking sequence is unlikely to have an effect.

The genome of Arabidopsis has relatively small introns and little intergenic material (Krysan et al. 1999), which makes insertional mutagenesis efficient. In the last decade, countless insertional lines have been made available to the Arabidopsis research community (Alonso et al. 2003). Without this effort, the functional analysis and study of genetic interactions within small gene families would have been strongly impeded. However, similar initiatives to create insertional stocks for other species seem infeasible. The complicated regeneration of transgenic plants from tissue cultures, long generation times, and short seed viability periods are some of the hurdles faced in other model species (e.g., poplar).

We examined the expression of all Arabidopsis KNOX genes in the stm-GK, knat1^bp−9, and knat7 mutants based on hypocotyl tissue samples (Table 5). Relative quantification was performed and the mathematical model from Pfaffl (2001) was also employed as a comparison. While KNAT1, KNAT2, KNAT4, and KNAT6 were slightly upregulated in the stm-GK mutant, no significantly different expression levels were observed for KNAT3, KNAT5, or KNAT7 (Table 5). Moreover, no significant differences were observed for all KNOX gene (except KNAT1 itself) expression in the knat1^bp−9 mutant compared to the wild type (Col-0), with p-values greater than 0.05. Interestingly, the expression of Arabidopsis KNOX genes (except for KNAT7) was significantly upregulated in the knat7 mutant. These results indicate that the expression study of KNOX in the Arabidopsis knox mutants can be observed and evaluated using established methods. Thus, expression studies of KNOX or other genes that are involved in or interact with KNOX genes can be further investigated using the same protocol.

Table 5.

Relative expression levels* and fold changes** of KNOX gene expression in stm-GK, knat1^bp−9, stm-GK, and knat1^bp−9 compared to the wild type (Col-0)

Lines	X	SD	p	Fold change (times)***	Fold Change (times)****	Expression
STM
Col-0	3.4507	0.3456
stm-GK	0.0046	0.0078	0.0017	− 745.70	− 6652.87	Downregulated
knat1bp−9	2.2576	0.2015	0.0559	− 1.53	− 1.83	Downregulated
knat7	14.2251	0.9993	0.0006	+ 4.12	+ 4.13	Upregulated
KNAT1
Col-0	1.3069	0.0915
stm-GK	1.9664	0.3979	0.0481	+ 1.50	+ 1.48	Upregulated
knat1bp−9	0.0054	0.0032	0.0008	− 242.84	− 286.84	Downregulated
knat7	0.9496	0.0157	0.0094	− 1.38	− 1.37	Downregulated
KNAT2
Col-0	2.2320	0.3455
stm-GK	3.2588	0.3866	0.0135	+ 1.46	+ 1.46	Upregulated
knat1bp−9	2.0057	0.4421	0.2627	− 1.11	− 1.12	Downregulated
knat7	7.9767	0.2630	0.0000	+ 3.57	+ 3.59	Upregulated
KNAT3
Col-0	0.0940	0.0211
Stm-GK	0.1067	0.0112	0.2125	+ 1.13	+ 1.15	Upregulated
knat1bp−9	0.1232	0.0548	0.2310	+ 1.31	+ 1.25	Upregulated
knat7	0.2537	0.0131	0.0005	+ 2.70	+ 2.74	Upregulated
KNAT4
Col-0	0.2234	0.0589
stm-GK	0.4102	0.0590	0.0089	+ 1.84	+ 1.87	Upregulated
knat1bp−9	0.3115	0.0189	0.0557	+ 1.39	+ 1.43	Upregulated
knat7	0.7637	0.0350	0.0003	+ 3.42	+ 3.51	Upregulated
KNAT5
Col-0	1.0151	0.2136
stm-GK	1.2794	0.3057	0.1471	+ 1.26	+ 1.26	Upregulated
knat1bp−9	1.4302	0.2708	0.0546	+ 1.41	+ 1.41	Upregulated
knat7	4.2369	0.1208	0.0001	+ 4.17	+ 4.24	Upregulated
KNAT6
Col-0	0.0134	0.0031
stm-GK	0.0257	0.0046	0.0114	+ 1.92	+ 1.94	Upregulated
knat1bp−9	0.0239	0.0071	0.0537	+ 1.78	+ 1.77	Upregulated
knat7	0.0557	0.0009	0.0004	+ 4.15	+ 4.24	Upregulated
KNAT7
Col-0	0.1866	0.0315
stm-GK	0.1385	0.0226	0.0527	− 1.35	− 1.24	Downregulated
knat1bp−9	0.1520	0.0165	0.0949	− 1.23	− 1.22	Downregulated
knat7	0.0013	0.0002	0.0047	− 139.93	− 139.40	Downregulated

X Mean value, SD standard deviation, p p value of statistical analysis using a t test compared to the wild type (Col)

*Relative concentration values were normalized to the ACTIN2 expression

**Calculated by dividing the mean of relative concentration values of knox mutants by the value for the wild type (downregulated or upregulated)

***Calculated using the formula in this work

****Calculated using Pfaffl’s (2001) formula. All data were calculated from three biological and three technical replicates.

Protocol overview and outlook

KNOX genes are families of homeobox genes that are involved in or regulate meristem function and many other biological functions in higher plants (Scofield and Murray 2006). In Arabidopsis, class I KNOX genes have been extensively studied. The history of these advances in research can be viewed as milestones in the literatures (Barton and Poethig 1993; Long et al. 1996; Chuck et al. 1996; Venglat et al. 2002; Mele et al. 2003; Dean et al. 2004; Hake et al. 2004; Scofield and Murray 2006; Belles-Boix et al. 2006; Hay and Tsiantis 2010; Scofield et al. 2013; Woerlen et al. 2017). KNOX homologs have also been explored in many plants based on functional studies of KNOX genes in Arabidopsis. However, class II KNOX genes are only scarcely described and functional data is largely lacking. As previously noted, studies of KNAT3 and KNAT7 have been reported (Truernit and Haseloff. 2007; Li et al. 2012; Qin et al. 2020).

Generating an established and reproducible protocol is very important for genetic research, especially for lengthy serial experiments. The present manuscript presents an expression study of the KNOX genes in Arabidopsis insertional mutants based on the initial works (i.e., from the screening of the homozygous mutant line to the final step of transcript analyses using qRT-PCR). The protocol flow chart is presented (Fig. 8). Moreover, issues and false results that might occur are also presented and discussed. This information provides useful insights for future genetic studies, especially those involving KNOX genes. Moreover, the information will also be useful for those in the early stages of working with genetic studies of Arabidopsis (i.e., those that lack experience). Serial experiments, such as the one presented in this protocol, must be combined with precise phenotypic observations especially of mutant plants to determine the specific functions of genes. Using an established method for working with knox mutant crossing (e.g., Weigel and Glazebrook 2006) will provide further information on the interactions of KNOX genes with other genes toward certain plant phenotypes. Since most KNOX proteins are transcription factors that regulate many functions in plant development, the study of downstream processes involving genes regulated by KNOX will be very interesting and open insights into the developmental biology and mechanisms of plants, especially Arabidopsis. Ultimately, this paper provides a fundamental work protocol that will open a variety of opportunities for further genetic studies of genes, especially those involving insertional mutants in Arabidopsis.

Fig. 8.

Protocol flow chart

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

The author designed, performed, and analyzed the experiments, as well as wrote the manuscript. The author got valuable guidances and advises from Dr. Urs Fischer (Department of Forest Botany and Tree Physiology, University of Goettingen) in research design, material selection, data analysis and interpretation. Author also got significant advices and financial support from Prof. Andrea Polle (The Head of Department of Forest Botany and Tree Physiology, University of Goettingen) through the poplar research group “Deutsche Forschungsgemeinschaft ‘Pappelgruppe’.

Funding

This work was funded by the Deutsche Forschungsgemeinschaft ‘Pappelgruppe’ (Andrea.Polle),

Declarations

Conflict of interest

The author declares no competing financial interests.