Two Iron-Responsive Promoter Elements Control Expression of FOX1 in Chlamydomonas reinhardtii

Abstract

FOX1 encodes an iron deficiency-induced ferroxidase involved in a high-affinity iron uptake system. Mutagenesis analysis of the FOX1 promoter identified two separate iron-responsive elements, FeRE1 (CACACG) and FeRE2 (CACGCG), between positions −87 and −82 and between positions −65 and −60, respectively, and both are needed for induced FOX1 expression under conditions of iron deficiency.

Iron and other transition metals are essential for virtually all forms of living organisms; however, iron overload can lead to the production of highly reactive hydroxyl radicals through the Fenton reaction. To manage this balancing act between deficiency and overload, a cell must have precise mechanisms to regulate its gene/protein expression and its iron uptake and use. How this regulation is achieved in animal systems and in yeast is well characterized, but almost nothing is known about how these mechanisms function in plants. It is important that we better understand this regulation in plants, since iron deficiency frequently limits both the growth of microalgae in the oceans (2) and that of agricultural crops (11). The former is important since oceanic microalgae are a major sink for atmospheric CO₂ and the latter because it leads to reduced yields and decreased nutritional quality of crops. In plants, two major strategies to acquire iron have evolved. Nongraminaceous plants use strategy I, a reduction strategy in which Fe³⁺ is solubilized in the rhizosphere by extruded protons, reduced to Fe²⁺ by an iron reductase, and taken up by an iron transporter. The other iron uptake strategy in plants, strategy II, is found in grasses (graminaceous monocots). Phytosiderophores are secreted into the rhizosphere, where they form stable Fe(III) chelates, which are transported into the root cells by specific transport systems (4, 20, 22). No iron-responsive promoter element (FeRE) that controls the expression of proteins induced by iron deficiency has been experimentally identified in a strategy I plant. However, in Arabidopsis thaliana, as many as 72 to 179 iron-regulated genes depend on the basic helix-loop-helix (bHLH) transcription factor FIT1 for their expression (3), and although the FeRE that FIT1 binds to has not been determined, bHLH factors are known to bind to the hexanucleotide E-box sequence CANNTG. A homolog of FIT1, FER, has also been identified in tomato (Lycopersicon esculentum) (18). In the strategy II plant barley, two FeREs have been identified in the IDS2 gene (15, 16). These two FeREs are IDE1 (ATCAAGCATGCTTCTTGC) and IDE2 (TTGAACGGCAAGTTTCACGCTGTCACT). The exact 5′ and 3′ borders of the cores of the FeREs in IDE1 and IDE2 have not been determined (IDE1 and IDE2 were defined by mutating the nucleotides of promoter regions needed for iron-regulated expression nine at a time), but when other iron-responsive genes from strategy II plants were searched for this sequence, several genes with similar motifs were found and a consensus sequence of CAAGCCTGCTTC(T/A)TGC was defined (16). Motifs fitting this consensus were also found in three A. thaliana genes (strategy I plant), but it has not been experimentally proven that these motifs act as FeREs in these genes. To better understand iron uptake in plants, we wanted to identify FeREs in a strategy I plant, and when studying plant processes, a valuable tool is often the use of less-complex photosynthetic organisms as model systems. One organism used to elucidate many processes in photosynthesis and other plant functions is the eukaryotic unicellular green alga Chlamydomonas reinhardtii (13). In C. reinhardtii, iron reductase and iron uptake activities are induced under iron-deficient conditions, and it is therefore proposed to use an iron uptake system similar to that of strategy I plants (9); this organism is especially well suited for use when studying metal-regulated gene expression, because since it requires only a few simple salts to grow, it is easy to control the exact metal concentrations in its growth medium. In the work presented here, we have analyzed the promoter of FOX1, which encodes an iron deficiency-induced ferroxidase involved in iron uptake (17). We have used C. reinhardtii strain CC425 (cw15 arg2), which was grown according to the method of Harris (12) in either liquid Tris-acetate-phosphate (TAP) medium in an orbital shaker (250 rpm, 25°C, and continuous light of 150 μmol m⁻² s⁻¹) or on solid TAP agar plates (22°C and 100 μmol m⁻² s⁻¹). The medium was prepared either with normal trace elements, giving it an iron concentration of 18 μM (iron sufficient) (+Fe), or with iron-free trace elements, giving it an iron concentration of 20 nM Fe (iron deficient) (−Fe). To analyze the FOX1 promoter for FeREs, we made a series of 5′ and 3′ deletions of the promoter and fused them to the reporter gene ARS (arylsulfatase) in either plasmid pJD54 or pJD100. In pJD54 the reporter gene is promoterless (6), and in pJD100 it is driven by a minimal promoter element derived from the C. reinhardtii β₂-tubulin gene (5). These constructs were transformed into C. reinhardtii by cotransformation with pArg7.8 (7, 14), and obtained transformants were screened for secreted arylsulfatase after 24 h on XSO₄-TAP plates with or without Fe (+/−Fe) (6) (the presence of a blue halo around the colony); for key constructs, we measured the transcription of ARS on three representative transformants by using real-time PCR under +/−Fe conditions. The results are summarized in Fig. 1, and the conclusion was that the FeRE(s) of FOX1 is located in the −103 to −41 region relative to the transcription start site. The sequences of all constructs were confirmed by DNA sequencing, and primers used to make them are listed in Table 1.

FIG. 1.

The FeRE of FOX1 is in the −103 to −41 promoter region. A series of nested deletions of the −1439 to +65 FOX1 promoter region were amplified by PCR, fused to the ARS reporter gene either in pJD54 (constructs pJF1439 to pJFD1) or in pJD100 (constructs pJF61R and pJF41R), and transformed into the arginine-requiring C. reinhardtii strain CC425 by cotransformation with pArg7.8. Arginine-independent colonies were transferred to +/−Fe XSO₄-TAP agar plates and incubated for 24 h to visualize arylsulfatase activity. The fraction of arylsulfatase-expressing colonies among the arginine-independent colonies 24 h after transfer is indicated in the ARS+/ARG+ column, with the cotransformation frequency shown in parentheses. The cotransformation frequency was determined with 16 randomly chosen transformants by PCR using one primer from the FOX1 promoter region and the other from ARS (for primers, see Table 1, Cotransformation 1). The relative increase in expression after transfer to −Fe conditions was determined by real-time PCR. The quotients of expression under −Fe and +Fe conditions were determined in triplicate with three randomly chosen transformants of each construct, and the expression was calculated based on the 2^−ΔΔCT method (19) using 18s rRNA amplification as an internal standard. For primers, see Table 1. nd, not determined.

TABLE 1.

PCR primers^a

Construct	Forward primer	Reverse primer	Sequence	Region amplified
pFOX1p		Fox13b	5′-TGTCCATGGGCGCTGGTGCTGTTA-3′	−1439 to +840
	Fox15a		5′-TTAAGCGTCAAGCCAACTCTGATTATG-3′
pJF1439		SalI-65R	5′-GAAGTCGACGGAAAGCAAGTGTCGCCA-3′	−1439 to +65
	KpnI-1439F		5′-TAAGGTACCAGCCAACTCTGATTAAG-3′
pJF103	KpnI-103F		5′-GCGGGTACCCAAGCTGCATCGTCGGCA-3′	−103 to +65
pJF77	KpnI-77F		5′-TATGGTACCGGCGCGGCTTGACACG-3′	−77 to +65
pJFD1	KpnI-1439F		5′-TAAGGTACCAGCCAACTCTGATTATG-3′	−1439 to +65 with the −103 to −78 region deleted
		104R	5′-AAGCCGCGCCCGATGGCGCCCGGTTA-3′
	77F		5′-GGCGCCATCGGGCGCGGCTTGACACG-3′
		SalI-65R	5′-GAAGTCGACGGAAAGCAAGTGTCGCCA-3′
Cotransformation 1		ArsR	5′-TTCTGAATGGCGTCCTGGTC-3′	FOX1-ARS2 junction
			See KpnI-1439F to KpnI-77F above
Cotransformation 2		ArsR	5′-TTCTGAATGGCGTCCTGGTC-3′	FOX1-ARS2 junction
	Fox94F		5′-TCGTCGGCACACGCCCAG-3′
RT-PCR ARS		ArsR1	5′-GTAGCGGATGTACTTGTGCAG-3′	RT-PCR of ARS2
	ArsF		5′-ATGGGTGCCCTCGCGGTGTTC-3′
RT-PCR control		18SrRNAR	5′-CCGTGTCAGGATTGGGTAATTT-3′	RT-PCR of 18S rRNA
	18SrRNAF		5′-TCAACTTTCGATGGTAGGATAGTG-3′
pJF41R	KpnI-103F		5′-GCGGGTACCCAAGCTGCATCGTCGGCA-3′	−103 to −41
		KpnI-41R	5′-AATGGTACCATCCCGGCCCGTCCCAAG-3′
pJF61R		KpnI-61R	5′-AATGGTACCGCGTGTCAAGCCGCGC-3′	−103 to −61

^aRT-PCR, real-time PCR.

To determine exactly which nucleotides in the −103 to −41 region are important for the regulation, we used a scanning mutagenesis analysis strategy, and for methodological reasons we analyzed the −103 to −77 region and the −76 to −41 region separately. To analyze the −103 to −77 region, nine constructs were generated by mutating the nucleotides in this region three by three in a −103 to +65 fragment and fusing these with pJD54 (Table 2, M103F to M79F). The mutations were made by using PCR with a forward primer, covering the region containing the desired mutation and a KpnI restriction site at the 5′ end, and the reverse primer, SalI-65R (the sequence complementary to +46 to +65 relative to the FOX1 transcription start site and a SalI restriction site). To better quantify the expression of arylsulfatase, the activity was measured in liquid cultures using the method described by de Hostos et al. (8). In Table 2, it can be seen that the mutations in three of these constructs (M88F, M85F, and M82F) render them noninducible, showing that the sequence from −88 to −80 (GCACACGCC) contains the core of a FeRE. In a second series of mutagenesis, these nine nucleotides were mutagenized one by one to each of three other possible nucleotides (Table 2, FM2 to FM28), and from the lack of inducible arylsulfatase activity in cells transformed with constructs FM5 to FM9, FM11 to FM15, FM17 to FM19, FM21, and FM22, it could be determined that the sequence from −87 to −82, CACACG, is the core motif of this FeRE; we designated this motif FeRE1. From the single-nucleotide mutations, it can also be seen that the substitutions of A to G in construct FM10, A to G in construct FM16, and G to T in FM20 did not result in a significant decrease of arylsulfatase activity, indicating that changing the sequence to these nucleotides will also give a functional FeRE. The consensus sequence of the FOX1 FeRE1 is therefore C(A/G)C(A/G)C(G/T). To analyze the −76 to −41 region, these nucleotides were mutated one to five at a time in the −103 to −41 fragment pJF41R (Table 2, constructs M73R to M41R) and fused to pJD100. To make the mutations, a reverse primer covering the region containing the desired mutation and with a KpnI restriction site added at the 5′ end was used in combination with a forward primer with a KpnI restriction site and the −103 to −86 region. Iron responsiveness could be measured for all constructs except for M60R, M61R, and M65R, and this shows that the sequence from −65 to −60 (CACGCG) is another FOX1 FeRE. This FeRE is not identical to FeRE1, but it fits the FeRE1 consensus sequence, and we designated this element FeRE2. It should, however, be noted that FeRE1 is still functional if the last nucleotide is changed from a G to a T, whereas doing the same substitution in FeRE2 abolishes iron regulation (constructs FM20 and M60R in Table 2). From the observation that with nonfunctional FeREs we receive a low level of expression under both +Fe and −Fe conditions, we can also draw the conclusion that these FeREs function as binding sites for activators during iron-deficient conditions and not for repressors under iron-sufficient conditions. If they had been repressor elements, with mutated FeREs, we would instead have seen an increase of expression under +Fe conditions.

TABLE 2.

Scanning mutagenesis of the FOX1 −103 to −41 region

Construct	Sequence	ARS activitya		No. of trans.b	Fe responsec
		−Fe	+Fe
	−103−77−41
	|||
pJF103	CAAGCTGCATCGTCGGCACACGCCCAGGCGCGGCTTGACACGCGCCTTGGGACGGGCCGGGAT	4.6 ± 1.7	0.5 ± 0.2	5	Yes
M103F	ACC------------------------------------------------------------	3.7 ± 2.0	0.4 ± 0.1	7	Yes
M100F	---TAG---------------------------------------------------------	4.4 ± 2.3	0.4 ± 0.2	7	Yes
M97F	------TAC------------------------------------------------------	2.1 ± 1.6	0.2 ± 0.1	10	Yes
M94F	---------GAT---------------------------------------------------	2.2 ± 0.9	0.2 ± 0.2	7	Yes
M91F	------------GAT------------------------------------------------	2.6 ± 1.3	0.1 ± 0.0	7	Yes
M88F	---------------TAC---------------------------------------------	0.2 ± 0.1	0.2 ± 0.1	7	No (63)
M85F	------------------ACA------------------------------------------	0.2 ± 0.2	0.2 ± 0.1	7	No (75)
M82F	---------------------TAA---------------------------------------	0.4 ± 0.2	0.3 ± 0.1	7	No (56)
M79F	------------------------ACT------------------------------------	2.0 ± 1.3	0.1 ± 0.1	7	Yes
FM1	--------------T------------------------------------------------	3.2 ± 1.4	0.2 ± 0.1	7	Yes
FM2	---------------T-----------------------------------------------	1.8 ± 0.6	0.3 ± 0.1	5	Yes
FM3	---------------C-----------------------------------------------	2.0 ± 1.0	0.3 ± 0.1	5	Yes
FM4	---------------A-----------------------------------------------	2.8 ± 1.4	0.2 ± 0.1	5	Yes
FM5	----------------A----------------------------------------------	0.1 ± 0.0	0.2 ± 0.1	5	No (63)
FM6	----------------T----------------------------------------------	0.2 ± 0.1	0.2 ± 0.2	5	No (69)
FM7	----------------G----------------------------------------------	0.3 ± 0.1	0.3 ± 0.1	4	No (75)
FM8	-----------------C---------------------------------------------	0.3 ± 0.2	0.3 ± 0.1	4	No (69)
FM9	-----------------T---------------------------------------------	0.3 ± 0.1	0.2 ± 0.1	5	No (56)
FM10	-----------------G---------------------------------------------	4.1 ± 0.7	0.5 ± 0.1	7	Yes
FM11	------------------A--------------------------------------------	0.1 ± 0.1	0.2 ± 0.1	4	No (50)
FM12	------------------T--------------------------------------------	0.3 ± 0.1	0.4 ± 0.1	6	No (81)
FM13	------------------G--------------------------------------------	0.4 ± 0.1	0.3 ± 0.1	5	No (63)
FM14	-------------------C-------------------------------------------	0.4 ± 0.1	0.2 ± 0.1	5	No (69)
FM15	-------------------T-------------------------------------------	0.5 ± 0.2	0.5 ± 0.2	4	No (63)
FM16	-------------------G-------------------------------------------	5.6 ± 1.8	0.3 ± 0.1	7	Yes
FM17	--------------------A------------------------------------------	0.1 ± 0.0	0.2 ± 0.1	4	No (56)
FM18	--------------------T------------------------------------------	0.1 ± 0.0	0.2 ± 0.0	4	No (75)
FM19	--------------------G------------------------------------------	0.5 ± 0.2	0.3 ± 0.1	4	No (63)
FM20	---------------------T-----------------------------------------	2.5 ± 1.4	0.2 ± 0.1	7	Yes
FM21	---------------------C-----------------------------------------	0.3 ± 0.1	0.2 ± 0.1	4	No (50)
FM22	---------------------A-----------------------------------------	0.3 ± 0.2	0.3 ± 0.1	4	No (69)
FM23	----------------------A----------------------------------------	1.5 ± 0.6	0.2 ± 0.1	6	Yes
FM24	----------------------T----------------------------------------	1.1 ± 0.5	0.2 ± 0.1	6	Yes
FM25	----------------------G----------------------------------------	2.1 ± 0.8	0.1 ± 0.1	6	Yes
FM26	-----------------------A---------------------------------------	2.1 ± 1.1	0.3 ± 0.1	6	Yes
FM27	-----------------------T---------------------------------------	1.4 ± 0.6	0.2 ± 0.0	6	Yes
FM28	-----------------------G---------------------------------------	4.0 ± 0.6	0.5 ± 0.3	6	Yes
FM29	------------------------A--------------------------------------	3.4 ± 1.5	0.5 ± 0.2	6	Yes
pJF41R	CAAGCTGCATCGTCGGCACACGCCCAGGCGCGGCTTGACACGCGCCTTGGGACGGGCCGGGAT	6.2 ± 4.0	0.9 ± 0.3	7	Yes
M73R	---------------------------TATA--------------------------------	4.1 ± 0.8	0.4 ± 0.3	6	Yes
M70R	-------------------------------TTA-----------------------------	3.9 ± 1.0	0.3 ± 0.2	6	Yes
M67R	----------------------------------GGT--------------------------	5.3 ± 3.3	0.3 ± 0.2	6	Yes
M66R	-------------------------------------C-------------------------	3.5 ± 1.2	0.4 ± 0.2	6	Yes
M65R	--------------------------------------A------------------------	0.2 ± 0.2	0.3 ± 0.1	4	No (50)
M61R	---------------------------------------CATA--------------------	0.3 ± 0.3	0.3 ± 0.1	4	No (50)
M60R	-------------------------------------------T-------------------	0.3 ± 0.1	0.3 ± 0.1	4	No (56)
M59R	--------------------------------------------A------------------	1.4 ± 1.3	0.3 ± 0.3	6	Yes
M56R	---------------------------------------------AGG---------------	1.0 ± 1.0	0.3 ± 0.2	6	Yes
M53R	-----------------------------------------------GTTT------------	5.5 ± 2.0	0.4 ± 0.2	6	Yes
M50R	---------------------------------------------------CAT---------	2.7 ± 0.9	0.2 ± 0.2	6	Yes
M47R	-----------------------------------------------------TTTA------	3.3 ± 2.1	0.2 ± 0.2	6	Yes
M44R	--------------------------------------------------------AATT---	4.2 ± 1.4	0.3 ± 0.1	6	Yes
M41R	----------------------------------------------------------TTTCG	4.4 ± 2.3	0.4 ± 0.3	6	Yes

^aExpressed as nmol p-nitrophenol min⁻¹ ·10⁻⁶ cells. The activities represent the averages for n independent arylsulfatase (ARS)-expressing transformants analyzed under both conditions.

^bNo. of transformants used to determine arylsulfatase activity.

^cFor constructs not giving iron-responsive expression, the cotransformation frequency (percent; in parentheses) was determined for 16 randomly picked transformants by PCR with one primer from the FOX1 promoter region and the other from the arylsulfatase gene (for primers, see Table 1, Cotransformation 2).

When searching for the FeRE1 consensus in the IDE1 and IDE2 elements in barley, the only exact match was the sequence CACGCT in IDE2, but since the T in this sequence is outside the IDE consensus motif and because of the poor similarity to IDE1 in this region, we doubt that these are homologous motifs. A more striking similarity to the IDE consensus is seen directly upstream of FeRE1; if a 1-nucleotide gap is inserted, 11 Chlamydomonas nucleotides are identical to the 16-nucleotide IDE consensus (Fig. 2). From the data in Table 2, we know that this region is not needed for iron-regulated expression of FOX1, but further investigations are needed to clarify if this sequence has a regulatory role under circumstances not tested in this study. There are two other systems to regulate gene expression under iron deficiency that have been described. In Saccharomyces cerevisiae, the transcription factors Aft1p and Aft2p activate the transcription of a set of more than 10 genes involved in the reductive and nonreductive uptake of iron into the cell (21), and they bind to a common FeRE motif with the consensus core (C/T)(A/G)CACCC(A/G). When the C. reinhardtii FeREs are compared with this consensus motif, it is seen that five out of six Chlamydomonas nucleotides are identical to the yeast consensus FeRE sequence (Fig. 2). We want to emphasize that this similarity is not enough to show that these two systems are homologous, but it is important to note that they are similar. We have searched the C. reinhardtii genome and expressed-sequence-tag database using the protein sequences of Aft1p and Aft2p, but we cannot find any Chlamydomonas homologues to these two transcription factors. Another aspect that argues against similar transcription factors is the fact that in yeast only one copy of the FeRE is needed, whereas in the Chlamydomonas FOX1 promoter two FeREs are needed to get iron-regulated gene expression. In both Arabidopsis and tomato, a putative bHLH transcription factor controls the induction of a number of genes at the onset of iron deficiency (3, 18). The nucleotide sequence that such transcription factors bind to is CANNTG, but although five of six nucleotides in the Chlamydomonas FeREs fit this motif (see Fig. 2), we do not believe these systems are homologous. The strongest argument against such homology is that the fifth nucleotide of the bHLH-binding site must be a T, whereas in Chlamydomonas, to get a functioning FeRE a C must be at this position (Table 2). Other arguments against similar regulatory systems in Chlamydomonas and Arabidopsis/tomato are the same as in the case with the yeast: only one copy of the bHLH motif is needed in Arabidopsis and tomato, and no FIT1 homolog can be found when searching the Chlamydomonas genome. During review of the manuscript, FeREs in another Chlamydomonas gene, ATX1, were identified (10), but the sequences are completely different from those of the FeREs identified in this work. When we search five other Chlamydomonas genes also shown to be iron responsive at the transcriptional level (1, 17) for FOX1 and ATX1 FeREs, the FeREs of FOX1 are found upstream of FER1, FEA1, and FEA2 and the ATX1 FeREs upstream of FTR1, FRE1, FEA1, and FEA2. In conclusion, the data presented here show that the FeREs in FOX1 are different from those in all other previously described iron-regulatory systems and that there are (at least) two classes of FeREs in Chlamydomonas.

FIG. 2.

Comparison of the FOX1 FeREs with FeREs in barley and yeast and with the E-box sequence that bHLH transcription factors bind to. Nucleotides with a black background are identical to FeRE1 or to the sequence upstream of FeRE1, and the gray background denotes nucleotide alternatives that also give a functional FOX1 FeRE1.